scholarly journals Multi-task weak supervision enables anatomically-resolved abnormality detection in whole-body FDG-PET/CT

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Sabri Eyuboglu ◽  
Geoffrey Angus ◽  
Bhavik N. Patel ◽  
Anuj Pareek ◽  
Guido Davidzon ◽  
...  

AbstractComputational decision support systems could provide clinical value in whole-body FDG-PET/CT workflows. However, limited availability of labeled data combined with the large size of PET/CT imaging exams make it challenging to apply existing supervised machine learning systems. Leveraging recent advancements in natural language processing, we describe a weak supervision framework that extracts imperfect, yet highly granular, regional abnormality labels from free-text radiology reports. Our framework automatically labels each region in a custom ontology of anatomical regions, providing a structured profile of the pathologies in each imaging exam. Using these generated labels, we then train an attention-based, multi-task CNN architecture to detect and estimate the location of abnormalities in whole-body scans. We demonstrate empirically that our multi-task representation is critical for strong performance on rare abnormalities with limited training data. The representation also contributes to more accurate mortality prediction from imaging data, suggesting the potential utility of our framework beyond abnormality detection and location estimation.

2020 ◽  
Author(s):  
Kenji Hirata ◽  
Osamu Manabe ◽  
Keiichi Magota ◽  
Sho Furuya ◽  
Tohru Shiga ◽  
...  

Abstract Background Radiology reports contribute not only to the particular patient, but also to constructing massive training dataset in the era of artificial intelligence (AI). The maximum standardized uptake value (SUVmax) is often described in daily radiology reports of FDG PET-CT. If SUVmax can be used as an identifier of lesion, that would greatly help AI interpret radiology reports. We aimed to clarify whether the lesion can be localized using SUVmax written in radiology reports.Methods The institutional review board approved this retrospective study. We investigated a total of 112 lesions from 30 FDG PET-CT images acquired with 3 different scanners. SUVmax was calculated from DICOM files based on the latest Quantitative Imaging Biomarkers Alliance (QIBA) publication. The voxels showing the given SUVmax were exhaustively searched in the whole-body images and counted. SUVmax was provided with 5 different degrees of precision: integer (e.g., 3), 1st decimal places (DP) (3.1), 2nd DP (3.14), 3rd DP (3.142), and 4th DP (3.1416). For instance, when SUVmax=3.14 was given, the voxels with 3.135≤SUVmax<3.145 were extracted. We also evaluated whether local maximum restriction could improve the identifying performance, where only the voxels showing the highest intensity within some neighborhood were considered. We defined that “identical detection” was achieved when only single voxel satisfied the criterion.Results A total of 112 lesions from 30 FDG PET-CT images were investigated. SUVmax ranged from 1.3 to 49.1 (median = 5.6, IQR = 5.2). Generally, when larger and more precise SUVmax values were given, fewer voxels satisfied the criterion. The local maximum restriction was very effective. When SUVmax was determined to 4 decimal places (e.g., 3.1416) and the local maximum restriction was applied, identical detection was achieved in 33.3% (lesions with SUVmax<2), 79.5% (2≤SUVmax<5), and 97.8% (5≤SUVmax) of lesions.Conclusions SUVmax of FDG PET-CT can be used as an identifier to localize the lesion if precise SUVmax is provided and local maximum restriction was applied, although the lesions showing SUVmax<2 were difficult to identify. The proposed method may have potential to make use of radiology reports retrospectively for constructing training datasets for AI.


2016 ◽  
Vol 85 (2) ◽  
pp. 459-465 ◽  
Author(s):  
Lino M. Sawicki ◽  
Johannes Grueneisen ◽  
Benedikt M. Schaarschmidt ◽  
Christian Buchbender ◽  
James Nagarajah ◽  
...  

2020 ◽  
Vol 50 (1) ◽  
pp. 249-254
Author(s):  
Miho Sasaki ◽  
Yuka Hotokezaka ◽  
Reiko Ideguchi ◽  
Masataka Uetani ◽  
Shuichi Fujita

AbstractMyositis ossificans (MO) is a benign soft-tissue lesion characterized by the heterotopic formation of the bone in skeletal muscles, usually due to trauma. MO is occasionally difficult to diagnose because of its clinical and radiological similarities with malignancy. We report a case of traumatic MO (TMO) in the masseter and brachial muscles of a 37-year-old man who presented with painless swelling in the left cheek and severe trismus. Due to the absence of a traumatic history at the first consultation and identification of a tumorous lesion in the left masseter muscle by magnetic resonance imaging (MRI), the lesion was suspected to be a malignant tumor. Subsequently, 18F-fluorodeoxyglucose positron-emission tomography/computed tomography (FDG-PET/CT) showed multiple regions of high FDG uptake across the whole body, suggestive of multiple metastases or other systemic diseases. However, intramuscular calcifications were also observed in the left masseter and brachial muscles, overlapping the areas with high FDG uptake. Moreover, multiple fractures were seen in the rib and lumbar spine, also overlapping the areas with high FDG uptake. Based on these imaging findings, along with a history of jet-ski trauma, TMO was suspected. The left cheek mass was surgically excised and histologically diagnosed as TMO. In this case report, FDG-PET/CT could detect multiple TMOs across the whole body. To the best of our knowledge, cases of multiple TMOs located far apart in different muscles are rare, and this may be the first report.


2013 ◽  
Vol 34 (6) ◽  
pp. 540-543 ◽  
Author(s):  
Kuruva Manohar ◽  
Anish Bhattacharya ◽  
Bhagwant R. Mittal
Keyword(s):  
Fdg Pet ◽  
Pet Ct ◽  
18F Fdg ◽  

Author(s):  
Olwen Westerland ◽  
◽  
Ashik Amlani ◽  
Christian Kelly-Morland ◽  
Michal Fraczek ◽  
...  

Abstract Purpose Comparative data on the impact of imaging on management is lacking for multiple myeloma. This study compared the diagnostic performance and impact on management of 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) and whole-body magnetic resonance imaging (WBMRI) in treatment-naive myeloma. Methods Forty-six patients undergoing 18F-FDG PET/CT and WBMRI were reviewed by a nuclear medicine physician and radiologist, respectively, for the presence of myeloma bone disease. Blinded clinical and imaging data were reviewed by two haematologists in consensus and management recorded following clinical data ± 18F-FDG PET/CT or WBMRI. Bone disease was defined using International Myeloma Working Group (IMWG) criteria and a clinical reference standard. Per-patient sensitivity for lesion detection was established. McNemar test compared management based on clinical assessment ± 18F-FDG PET/CT or WBMRI. Results Sensitivity for bone lesions was 69.6% (32/46) for 18F-FDG PET/CT (54.3% (25/46) for PET component alone) and 91.3% (42/46) for WBMRI. 27/46 (58.7%) of cases were concordant. In 19/46 patients (41.3%) WBMRI detected more focal bone lesions than 18F-FDG PET/CT. Based on clinical data alone, 32/46 (69.6%) patients would have been treated. Addition of 18F-FDG PET/CT to clinical data increased this to 40/46 (87.0%) patients (p = 0.02); and WBMRI to clinical data to 43/46 (93.5%) patients (p = 0.002). The difference in treatment decisions was not statistically significant between 18F-FDG PET/CT and WBMRI (p = 0.08). Conclusion Compared to 18F-FDG PET/CT, WBMRI had a higher per patient sensitivity for bone disease. However, treatment decisions were not statistically different and either modality would be appropriate in initial staging, depending on local availability and expertise.


2021 ◽  
Vol 5 (1) ◽  
pp. 1151-1160
Author(s):  
A.S. Lukashevich ◽  

Purpose. The purpose of the article is to evaluate the diagnostic significance of positron emission tomography / computed tomography with 18F -fluorodeoxyglucose (18F -FDG PET/CT) for the diagnosis of prosthetic endocarditis. Methods of research. The study included 82 patients with suspected prosthetic endocarditis in accordance with the criteria proposed by Duke University [1-5]. The patients received hospital treatment at the State Institution RSPC "Cardiology" from January 2016 to March 2021. The study was of a prospective, non-randomized, single-center cohort design. The duration of the monitor period was 12 months from the moment of patients’ inclusion in the study. Whole-body positron emission tomography / computed tomography (PET/CT) examinations were performed in 82 patients. 27 patients were selected for surgical treatment. Conservative treatment group included 16 patients. 27 patients were selected into the observation group, they were suspected to have prosthetic heart valve infection in the primary referral and underwent PET/CT scanning, according to which the diagnosis of prosthetic endocarditis was excluded. The event under the study did not develop in this group during the year of observation. Results and conclusion. The history of infective endocarditis was not statistically significant and did not increase the risk of developing prosthetic endocarditis in the sample presented. The Duke criteria are less reliable in establishing the diagnosis of prosthetic endocarditis. The median number of days from the date of the first prosthesis implantation to the onset of prosthetic endocarditis was about 4 years. This study revealed that the development of the infectious process in the area of the prosthesis was noted in a more distant postoperative period compared to literature data. Histological confirmation of infection was noted in 100% (27 patients) of cases in reoperated patients. The presence of a more formidable complication such as valve ring abscess located mainly in the projection of the aortic valve ring was quite common in both groups. Presepsin and Interleukin-6 have a statistically significant (U = 394,50 p = 0,01 and U = 94,50 p = 0.004) value in the prognosis of prosthetic endocarditis. Considering the data obtained from ROC analysis, it can be said that the cut-off point at which it is possible to diagnose prosthetic endocarditis based on PETCT is 2.85. The presented methods for the interpretation of whole-body FDG-PET/CT images of patients with suspected infectious complications after cardiac surgery, as well as with the presence of prosthetic endocarditis, show high sensitivity and specificity.


2012 ◽  
Vol 152 (4) ◽  
pp. 334-338 ◽  
Author(s):  
C. Etard ◽  
D. Celier ◽  
P. Roch ◽  
B. Aubert

Author(s):  
Almalki Yassir

Abstract Background Leydig cell tumors (LCTs) represent the most common form of stromal tumors. We reported the 18F-fluorodeoxyglucose (FDG) positron emission tomography (PET)/computed tomography (CT) findings of a patient with testicular LCT. Case presentation A 50-year-old man with a history of end-stage renal disease and renal transplantation 19 years ago. One year earlier, he started to have a chronic rejection. During the investigation to determine the cause of chronic rejection, a suspicious lesion in the graft with a collection around it was seen on ultrasound (US) images, raising the possibility of post-transplant lymphoproliferative disorder (PTLD). The patient was referred for further evaluation by whole body 18F-FDG PET/CT imaging. The image finding revealed an incidental hypermetabolic focal lesion in the right testicle—no other specific findings in the remaining parts of the body nor definitive FDG avid lymphadenopathy to suggest PTLD. Testicular US was requested and showed a well-defined right-sided heterogeneous hypoechoic intratesticular focal mass at the upper pole of the right testis with significant internal vascularity on the color Doppler imaging. The patient underwent a right radical orchidectomy, and the tumor was pathologically confirmed as an LCT. Conclusion In our case, 18F-FDG-PET/CT has been helpful in incidentally detecting this rare testicular tumor in a patient with suspected PTLD.


Sign in / Sign up

Export Citation Format

Share Document