Prostate and dominant intraprostatic lesion segmentation on PET/CT using cascaded regional-net

2021 ◽  
Author(s):  
Luke A Matkovic ◽  
Tonghe Wang ◽  
Yang Lei ◽  
Oladunni O Akin-Akintayo ◽  
Olayinka A Abiodun Ojo ◽  
...  
Keyword(s):  
Radiation Therapy ◽  
Cross Validation ◽  
Ct Images ◽  
Second Step ◽  
Prostate Segmentation ◽  
Positron Emission ◽  
Pet Ct ◽  
The Mean ◽  
And Training ◽  
Fold Cross Validation

Abstract Focal dose boost to dominant intraprostatic lesions (DILs) has recently been proposed for prostate radiation therapy. Accurate and fast delineation of the prostate and DILs is thus required during treatment planning. We propose a learning-based method using positron emission tomography (PET)/computed tomography (CT) images to automatically segment the prostate and its DILs. To enable end-to-end segmentation, a deep learning-based method, called cascaded regional-Net, is utilized. The first network, referred to as dual attention network (DAN), is used to segment the prostate via extracting comprehensive features from both PET and CT images. A second network, referred to as mask scoring regional convolutional neural network (MSR-CNN), is used to segment the DILs from the PET and CT within the prostate region. Scoring strategy is used to diminish the misclassification of the DILs. For DIL segmentation, the proposed cascaded regional-Net uses two steps to remove normal tissue regions, with the first step cropping images based on prostate segmentation and the second step using MSR-CNN to further locate the DILs. The binary masks of DILs and prostates of testing patients are generated from PET/CT by the trained network. To evaluate the proposed method, we retrospectively investigated 49 PET/CT datasets. On each dataset, the prostate and DILs were delineated by physicians and set as the ground truths and training targets. The proposed method was trained and evaluated using a five-fold cross-validation and a hold-out test. The mean surface distance and DSC values were 0.666±0.696mm and 0.932±0.059 for the prostate and 1.209±1.954mm and 0.757±0.241 for the DILs among all 49 patients. The proposed method has demonstrated great potential for improving the efficiency and reducing the observer variability of prostate and DIL contouring for DIL focal boost prostate radiation therapy.

scholarly journals Convolutional Neural Networks Promising in Lung Cancer T-Parameter Assessment on Baseline FDG-PET/CT

2018 ◽  
Vol 2018 ◽  
pp. 1-6 ◽  
Author(s):  
Margarita Kirienko ◽  
Martina Sollini ◽  
Giorgia Silvestri ◽  
Serena Mognetti ◽  
Emanuele Voulaz ◽  
...  
Keyword(s):  
Lung Cancer ◽  
Fdg Pet ◽  
Area Under The Curve ◽  
Ct Images ◽  
Tnm System ◽  
Positron Emission ◽  
Pet Ct ◽  
Test Sets ◽  
Fdg Pet Ct

Aim. To develop an algorithm, based on convolutional neural network (CNN), for the classification of lung cancer lesions as T1-T2 or T3-T4 on staging fluorodeoxyglucose positron emission tomography (FDG-PET)/CT images. Methods. We retrospectively selected a cohort of 472 patients (divided in the training, validation, and test sets) submitted to staging FDG-PET/CT within 60 days before biopsy or surgery. TNM system seventh edition was used as reference. Postprocessing was performed to generate an adequate dataset. The input of CNNs was a bounding box on both PET and CT images, cropped around the lesion centre. The results were classified as Correct (concordance between reference and prediction) and Incorrect (discordance between reference and prediction). Accuracy (Correct/[Correct + Incorrect]), recall (Correctly predicted T3-T4/[all T3-T4]), and specificity (Correctly predicted T1-T2/[all T1-T2]), as commonly defined in deep learning models, were used to evaluate CNN performance. The area under the curve (AUC) was calculated for the final model. Results. The algorithm, composed of two networks (a “feature extractor” and a “classifier”), developed and tested achieved an accuracy, recall, specificity, and AUC of 87%, 69%, 69%, and 0.83; 86%, 77%, 70%, and 0.73; and 90%, 47%, 67%, and 0.68 in the training, validation, and test sets, respectively. Conclusion. We obtained proof of concept that CNNs can be used as a tool to assist in the staging of patients affected by lung cancer.

Prognostic risk factors to predict final pathohistology in patients undergoing radical salvage prostatectomy for locally recurrent prostate cancer after radiation therapy

2007 ◽  
Vol 25 (18_suppl) ◽  
pp. 15544-15544
Author(s):  
D. Pfister ◽  
C. Ohlmann ◽  
D. Sahi ◽  
U. Engelmann ◽  
A. Heidenreich
Keyword(s):  
Prostate Cancer ◽  
Multivariate Analysis ◽  
Radiation Therapy ◽  
Gleason Score ◽  
Curative Intent ◽  
Ldr Brachytherapy ◽  
Pet Ct ◽  
The Mean ◽  
Locally Recurrent ◽  
Biopsy Gleason Score

15544 Background: Radical salvage prostatectomy (sRPE) represents one local secondary treatment option with curative intent in patients failing radiation therapy for localized prostate cancer (PCA). Currently, there are very few studies correlating preoperative clinical and pathohistological variables with final pathohistology of sRPE specimens. It was the purpose of our study to identify prognosticators predicting organ confined and locally advanced PCA. Methods: 45 patients with biopsy-proven locally recurrent PCA underwent sRPE and extended pelvic lymphadenectomy (epLA) via a retropubic approach. Preoperative PSA, PSA doubling time, PSA prior to initial radiation therapy, biopsy Gleason score, number of positive biopsies, cT stage, 11choline PET/CT findings, type of radiation therapy, neoadjuvant androgen deprivation were correlated with the pathohistological stage by uni- and multivariate analysis. Results: A total of 45 patients underwent sRPE and epLA; 16 (35.5%), 12 (26.6%) and 17 (37.8%) patients had undergone external beam radiation (EBRT), HDR and LDR brachytherapy, resp. The mean preop. serum PSA was 7.8 (2–24) ng/ml; mean biopsy Gleason score was 5.6 (4–9). We did not encounter significant intraoperative compliations, the mean blood loss was 490 (200–950) ml. A mean of 19 (10 - 32) lymph nodes were removed. Pathohistology showed stage pT1–2pN0 in 27 (60%), stage pT3a/b and pTxpN1 PCA in 9 (20%) and 9 (20%) of patients, respectively. Positive surgical margins were identified in 5 (11%) patients. By multivariate analysis the parameters significantly associated with organ confined PCA sRPE are PSADT > 12 months, = 50% positive biopsy cores, biopsy Gleason score = 7 and previous LDR brachytherapy (pT1–2pN0R0 in all men); pre-radiation and preoperative PSA, PET/CT findings had no significant impact with final pTpN-stage. Conclusions: SRPE can be performed with a low morbidity in biopsy proven locally recurrent PCA after radiotherapy. The identified prognostic parameters will help to select patients most suitable for a local secondary surgical approach with curative intent. Especially in patients with local relapse following LDR brachytherapy sRPE represents a valuable treatment option. No significant financial relationships to disclose.

scholarly journals Impression Classification of Endek (Balinese Fabric) Image Using K-Nearest Neighbors Method

2018 ◽  
pp. 213-220 ◽  
Author(s):  
Gede Aditra Pradnyana ◽  
I Komang Agus Suryantara ◽  
I Gede Mahendra Darmawiguna
Keyword(s):  
Cross Validation ◽  
Nearest Neighbors ◽  
K Nearest Neighbors ◽  
K Value ◽  
Training Samples ◽  
And Training ◽  
Validation Testing ◽  
Fold Cross Validation ◽  
Learning Data

An impression can be interpreted as a psychological feeling toward a product and it plays an important role in decision making. Therefore, the understanding of the data in the domain of impressions will be very useful. This research had the objective of knowing the performance of K-Nearest Neighbors method to classify endek image impression using K-Fold Cross Validation method. The images were taken from 3 locations, namely CV. Artha Dharma, Agung Bali Collection, and Pengrajin Sri Rejeki. To get the image impression was done by consulting with an endek expert named Dr. D.A Tirta Ray, M.Si. The process of data mining was done by using K-Nearest Neighbors Method which was a classification method to a set of data based on learning data that had been classified previously and to classify new objects based on attributes and training samples. K-Fold Cross Validation testing obtained accuracy of 91% with K value in K-Nearest Neighbors of 3, 4, 7, 8.

scholarly journals CV-α: designing validations sets to increase the precision and enable multiple comparison tests in genomic prediction

2020 ◽  
Author(s):  
Rafael Massahiro Yassue ◽  
José Felipe Gonzaga Sabadin ◽  
Giovanni Galli ◽  
Filipe Couto Alves ◽  
Roberto Fritsche-Neto
Keyword(s):  
Genomic Prediction ◽  
Cross Validation ◽  
Prediction Models ◽  
Mean Squared Error ◽  
Predictive Ability ◽  
Proof Of Concept ◽  
Squared Error ◽  
High Effect ◽  
The Mean ◽  
Fold Cross Validation

AbstractUsually, the comparison among genomic prediction models is based on validation schemes as Repeated Random Subsampling (RRS) or K-fold cross-validation. Nevertheless, the design of training and validation sets has a high effect on the way and subjectiveness that we compare models. Those procedures cited above have an overlap across replicates that might cause an overestimated estimate and lack of residuals independence due to resampling issues and might cause less accurate results. Furthermore, posthoc tests, such as ANOVA, are not recommended due to assumption unfulfilled regarding residuals independence. Thus, we propose a new way to sample observations to build training and validation sets based on cross-validation alpha-based design (CV-α). The CV-α was meant to create several scenarios of validation (replicates x folds), regardless of the number of treatments. Using CV-α, the number of genotypes in the same fold across replicates was much lower than K-fold, indicating higher residual independence. Therefore, based on the CV-α results, as proof of concept, via ANOVA, we could compare the proposed methodology to RRS and K-fold, applying four genomic prediction models with a simulated and real dataset. Concerning the predictive ability and bias, all validation methods showed similar performance. However, regarding the mean squared error and coefficient of variation, the CV-α method presented the best performance under the evaluated scenarios. Moreover, as it has no additional cost nor complexity, it is more reliable and allows the use of non-subjective methods to compare models and factors. Therefore, CV-α can be considered a more precise validation methodology for model selection.

scholarly journals A Prospective Observational Study to Compare the Effect of Adaptive Radiation Therapy in Head and Neck Cancer Patients Treated with Helical Tomotherapy – Bangalore, Karnataka

2021 ◽  
Vol 8 (41) ◽  
pp. 3578-3583
Author(s):  
Somorat Bhattacharjee ◽  
Sunil R.A. ◽  
Pichandi A. ◽  
Muthuselvi C.A. ◽  
Souradeep Bhattacharjee
Keyword(s):  
Spinal Cord ◽  
Radiation Therapy ◽  
Head And Neck Cancer ◽  
Parotid Gland ◽  
Ct Scan ◽  
Cancer Patients ◽  
Adaptive Radiation ◽  
Helical Tomotherapy ◽  
Pet Ct ◽  
The Mean

BACKGROUND Establishing the usefulness of adaptive radiotherapy in our setting with limited data might help to ensure better conformity and reduce treatment related morbidity. Hence we conducted this study to elicit the benefit of adaptive radiotherapy with helical tomotherapy. METHODS This is a prospective study conducted among 25 head and neck cancer patients undergoing radiotherapy with helical tomotherapy. All patients underwent initial radiation therapy treatment planning simulation positron emission tomography computed tomography (PET CT/ CT scan) [CT-1], followed by repeat PET CT/ CT scan at 4th - 5th week of radiotherapy [CT-2]. Planning for full intended dose [66 Gy - 70 Gy] was done on both the scans, keeping the radiation therapy planning parameters same. Changes in the volume of the clinical target volumes (CTV), changes in the volume and dose to spinal cord, bilateral parotids, and mandible were compared. A p - value of < 0.05 was considered for statistical significance. RESULTS A significant reduction in the volumes of tumour - CTV-1 [CT-1 v/s CT-2: 166.82 cc v/s. 150.63 cc] and of lymph nodal region - CTV-2 [CT-1 v/s CT-2: 260.29 cc v/s 228.00 cc], contra lateral parotid gland [CT-1 v/s CT-2: 33.00 cc v/s 18.72 cc] were observed (P < 0.05). The mean doses received by contra lateral parotid gland [CT-1 v/s. CT-2: 23.14 Gy v/s 21.26 Gy] were significantly lesser in the CT2 scans (P < 0.05). The mean maximum doses were also significantly lesser to the mandible and spinal cord i.e., CT-1 v/s. CT-2: 68.528 Gy v/s 67.39 Gy and 39.45 Gy v/s. 37.33 Gy respectively (P < 0.05). A significant reduction in standardised uptake value (SUV), values of the primary tumour and involved lymph nodes was observed between CT-1 and CT-2. CONCLUSIONS During 4th to 5th week of radiation therapy, significant reductions in the CTVs and in dose to OARs were noted. Thus, we recommend at least one re-simulation scan and re-planning during radiation therapy, irrespective of the type of technique of radiation therapy. KEYWORDS Adaptive Radiation Therapy, IMRT, Tomotherapy

scholarly journals Convolutional Neural Networks for automatic image quality control and EARL compliance of PET images

2021 ◽  
Author(s):  
Elisabeth Pfaehler ◽  
Daniela Euba ◽  
Andreas Rinscheid ◽  
Otto S. Hoekstra ◽  
Josee Zijlstra ◽  
...  
Keyword(s):  
Machine Learning ◽  
Cross Validation ◽  
Training Data ◽  
Independent Dataset ◽  
Pet Ct ◽  
Image Quality Control ◽  
The Cross ◽  
The Impact ◽  
Pet Scanners ◽  
Fold Cross Validation

Abstract Background Machine learning studies require a large number of images often obtained on different PET scanners. When merging these images, the use of harmonized images following EARL-standards is essential. However, when including retrospective images, EARL accreditation might not have been in place. The aim of this study was to develop a convolutional neural network (CNN) that can identify retrospectively if an image is EARL compliant and if it is meeting older or newer EARL-standards. Materials and Methods 96 PET images acquired on three PET/CT systems were included in the study. All images were reconstructed with the locally clinically preferred, EARL1, and EARL2 compliant reconstruction protocols. After image pre-processing, one CNN was trained to separate clinical and EARL compliant reconstructions. A second CNN was optimized to identify EARL1 and EARL2 compliant images. The accuracy of both CNNs was assessed using 5-fold cross validation. The CNNs were validated on 24 images acquired on a PET scanner not included in the training data. To assess the impact of image noise on the CNN decision, the 24 images were reconstructed with different scan durations. Results In the cross-validation, the first CNN classified all images correctly. When identifying EARL1 and EARL2 compliant images, the second CNN identified 100% EARL1 compliant and 85% EARL2 compliant images correctly. The accuracy in the independent dataset was comparable to the cross-validation accuracy. The scan duration had almost no impact on the results. Conclusion The two CNNs trained in this study can be used to retrospectively include images in a multi-center setting by e.g. adding additional smoothing. This method is especially important for machine learning studies where the harmonization of images from different PET systems is essential.

scholarly journals Improving Prediction Accuracy Using Multi-allelic Haplotype Prediction and Training Population Optimization in Wheat

2020 ◽  
Vol 10 (7) ◽  
pp. 2265-2273 ◽  
Author(s):  
Ahmad H. Sallam ◽  
Emily Conley ◽  
Dzianis Prakapenka ◽  
Yang Da ◽  
James A. Anderson
Keyword(s):  
Population Structure ◽  
Protein Content ◽  
Prediction Accuracy ◽  
Cross Validation ◽  
Predictive Ability ◽  
Training Population ◽  
Percentage Points ◽  
And Training ◽  
Fold Cross Validation ◽  
Single Snps

The use of haplotypes may improve the accuracy of genomic prediction over single SNPs because haplotypes can better capture linkage disequilibrium and genomic similarity in different lines and may capture local high-order allelic interactions. Additionally, prediction accuracy could be improved by portraying population structure in the calibration set. A set of 383 advanced lines and cultivars that represent the diversity of the University of Minnesota wheat breeding program was phenotyped for yield, test weight, and protein content and genotyped using the Illumina 90K SNP Assay. Population structure was confirmed using single SNPs. Haplotype blocks of 5, 10, 15, and 20 adjacent markers were constructed for all chromosomes. A multi-allelic haplotype prediction algorithm was implemented and compared with single SNPs using both k-fold cross validation and stratified sampling optimization. After confirming population structure, the stratified sampling improved the predictive ability compared with k-fold cross validation for yield and protein content, but reduced the predictive ability for test weight. In all cases, haplotype predictions outperformed single SNPs. Haplotypes of 15 adjacent markers showed the best improvement in accuracy for all traits; however, this was more pronounced in yield and protein content. The combined use of haplotypes of 15 adjacent markers and training population optimization significantly improved the predictive ability for yield and protein content by 14.3 (four percentage points) and 16.8% (seven percentage points), respectively, compared with using single SNPs and k-fold cross validation. These results emphasize the effectiveness of using haplotypes in genomic selection to increase genetic gain in self-fertilized crops.

Use of FDG-PET Imaging for Patients with Disseminated Cancer of the Appendix

2010 ◽  
Vol 76 (12) ◽  
pp. 1338-1344 ◽  
Author(s):  
Payam Rohani ◽  
Stephen D. Scotti ◽  
Perry Shen ◽  
John H. Stewart ◽  
Gregory B. Russell ◽  
...  
Keyword(s):  
Pet Imaging ◽  
Fdg Pet ◽  
Peritoneal Dissemination ◽  
Low Grade ◽  
High Grade ◽  
Positron Emission ◽  
Pet Ct ◽  
The Mean ◽  
Ct Data ◽  
Abdominal Quadrant

The goal of this study is to evaluate the use of positron emission tomography (PET) in evaluation of patients with peritoneal dissemination of carcinoma of appendiceal origin (PDA). Thirty-three patients with PDA, who had preoperative PET or PET/CT imaging, were analyzed. Using operative, pathology, and PET ± CT data, presence or absence of disease in each abdominal quadrant was noted and the use of 18fluoro-deoxy-glucose (FDG) PET for each quadrant was evaluated. The mean age was 52, and there were 17 males; 58 per cent had low-grade lesions. PET was positive in only 35 per cent of cases overall (30 and 41% sensitivity for low-grade and high-grade, respectively). PET without CT sensitivity for low-grade and high-grade lesions was 21 and 8 per cent, respectively. PET imaging has limited use for patients with PDA. We do not recommend the use of FDG-PET for patients with PDA from cancer of the appendix.

18-Fluoro-Deoxy-Glucose Positron Emission Tomography With Computed Tomography–Based Gross Tumor Volume Estimation and Validation With Magnetic Resonance Imaging for Locally Advanced Cervical Cancers

2012 ◽  
Vol 22 (6) ◽  
pp. 1031-1036 ◽  
Author(s):  
Maheshkumar N. Upasani ◽  
Umesh M. Mahantshetty ◽  
Venkatesh Rangarajan ◽  
Nilendu Purandare ◽  
Nikhil Merchant ◽  
...  
Keyword(s):  
Primary Tumor ◽  
Fdg Pet ◽  
Tumor Volume ◽  
Volume Estimation ◽  
Emission Tomography ◽  
Resonance Imaging ◽  
Positron Emission ◽  
Pet Ct ◽  
The Mean ◽  
Fdg Pet Ct

PurposeAnatomy and morphology–based imaging is routinely used for radiotherapy purpose to deliver precision treatment. There is an interest in using information from functional imaging for conformal radiation therapy planning. These functional imaging techniques need to be validated rigorously before their routine use. We attempted to evaluate and validate the use of 18-fluoro-deoxy-glucose positron emission tomography with computed tomography (18FDG PET-CT) on primary tumor of the cervical carcinoma, with an aim of arriving at a cutoff maximum standardized uptake value (SUVmax) at which the tumor volume correlates best with magnetic resonance imaging (MRI). This observational study was a part of an ethics committee–approved study evaluating pretreatment MRI and FDG PET-CT.Materials and MethodsPatients’ biopsy-proven cervical carcinomas (stages IIB and IIIB) were included in this study and underwent pretreatment MRI and FDG PET-CT as per institutional protocol. Volumes of the disease at the cervix on the MR image were calculated. Volumes at the FDG PET-CT scan at different percentages of SUVmax were auto contoured. Volume at MRI was correlated with each different percentage cutoff of the SUVmax.ResultsData of 74 patients were available for the study. The mean (SD) SUVmax of the primary tumor was 15.7 (7.0). The mean MRI volume correlates significantly (P < 0.001) with 30% and 35% of SUVmax values with good correlation according to the Pearson bivariate correlation (r = 0.79 each). The mean difference between MRI and PET volumes was least with 30% SUVmax.Conclusions18FDG PET-CT SUV-based primary tumor volume estimation at 30% to 35% of SUVmax values correlates significantly with the criterion standard MR volumes for primary cervical tumor with squamous histology in our population.

Sign in / Sign up

Export Citation Format

Share Document