P28 Weakly supervised deep learning on CT scans predicts survival from chronic pulmonary aspergillosis

2018 ◽  
Author(s):  
M Zusag ◽  
R Angheleanu ◽  
H Norhan ◽  
L Nwankwo ◽  
J Periselneris ◽  
...  
Keyword(s):  
Deep Learning ◽  
Ct Scans ◽  
Pulmonary Aspergillosis ◽  
Chronic Pulmonary Aspergillosis ◽  
Weakly Supervised

scholarly journals COVID-19 Diagnosis from Chest CT Scans: A Weakly Supervised CNN-LSTM Approach

AI ◽  
2021 ◽  
Vol 2 (3) ◽  
pp. 330-341
Author(s):  
Mustafa Kara ◽  
Zeynep Öztürk ◽  
Sergin Akpek ◽  
Ayşegül Turupcu
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Convolutional Neural Network ◽  
Short Term Memory ◽  
High Sensitivity ◽  
High Specificity ◽  
Chest Ct ◽  
Ct Scans ◽  
Imaging Data ◽  
Weakly Supervised

Advancements in deep learning and availability of medical imaging data have led to the use of CNN-based architectures in disease diagnostic assisted systems. In spite of the abundant use of reverse transcription-polymerase chain reaction-based tests in COVID-19 diagnosis, CT images offer an applicable supplement with their high sensitivity rates. Here, we study the classification of COVID-19 pneumonia and non-COVID-19 pneumonia in chest CT scans using efficient deep learning methods to be readily implemented by any hospital. We report our deep network framework design that encompasses Convolutional Neural Networks and bidirectional Long Short Term Memory architectures. Our study achieved high specificity (COVID-19 pneumonia: 98.3%, non-COVID-19 pneumonia: 96.2% Healthy: 89.3%) and high sensitivity (COVID-19 pneumonia: 84.0%, non-COVID-19 pneumonia: 93.9% Healthy: 94.9%) in classifying COVID-19 pneumonia, non-COVID-19 pneumonia and healthy patients. Next, we provide visual explanations for the Convolutional Neural Network predictions with gradient-weighted class activation mapping (Grad-CAM). The results provided a model explainability by showing that Ground Glass Opacities, indicators of COVID-19 pneumonia disease, were captured by our convolutional neural network. Finally, we have implemented our approach in three hospitals proving its compatibility and efficiency.

scholarly journals Classification of Multiple Diseases on Body CT Scans Using Weakly Supervised Deep Learning

2021 ◽  
Author(s):  
Fakrul Islam Tushar ◽  
Vincent M. D'Anniballe ◽  
Rui Hou ◽  
Maciej A. Mazurowski ◽  
Wanyi Fu ◽  
...  
Keyword(s):  
Deep Learning ◽  
Ct Scans ◽  
Weakly Supervised ◽  
Body Ct

scholarly journals COVID-19 Diagnosis from chest CT Scans: A Weakly Supervised CNN-LSTM Approach

2021 ◽  
Author(s):  
Mustafa Kara ◽  
Zeynep Öztürk ◽  
Sergin Akpek ◽  
Ayşegül Turupcu
Keyword(s):  
Deep Learning ◽  
Short Term Memory ◽  
High Sensitivity ◽  
High Specificity ◽  
Chest Ct ◽  
Ct Scans ◽  
Imaging Data ◽  
Framework Design ◽  
Weakly Supervised

Advancements in deep learning and availability of medical imaging data have led to use of CNN based architectures in disease diagnostic assisted systems. In spite of the abundant use of reverse transcription-polymerase chain reaction (RT-PCR) based tests in COVID-19 diagnosis, CT images offer an applicable supplement with its high sensitivity rates. Here, we study classification of COVID-19 pneumonia (CP) and non-COVID-19 pneumonia (NCP) in chest CT scans using efficient deep learning methods to be readily implemented by any hospital. We report our deep network framework design that encompasses Convolutional Neural Networks (CNNs) and bidirectional Long Short Term Memory (biLSTM) architectures. Our study achieved high specificity (CP: 98.3%, NCP: 96.2% Healthy: 89.3%) and high sensitivity (CP: 84.0%, NCP: 93.9% Healthy: 94.9%) in classifying COVID-19 pneumonia, non-COVID-19 pneumonia and healthy patients. Next, we provide visual explanations for the CNN predictions with gradient-weighted class activation mapping (Grad-CAM). The results provided a model explainability by showing that Ground Glass Opacities (GGO), indicators of COVID-19 pneumonia disease, were captured by our CNN network. Finally, we have implemented our approach in three hospitals proving its compatibility and efficiency.

PO-1718: Deep learning based auto-contouring of planning CT scans for rectal cancer

2020 ◽  
Vol 152 ◽  
pp. S949
Author(s):  
L. Bokhorst ◽  
M.H.F. Savenije ◽  
M.P.W. Intven ◽  
C.A.T. Van den Berg
Keyword(s):  
Rectal Cancer ◽  
Deep Learning ◽  
Ct Scans ◽  
Planning Ct

scholarly journals Misdiagnosis of chronic pulmonary aspergillosis as pulmonary tuberculosis at a tertiary care center in Uganda: a case series

2021 ◽  
Vol 15 (1) ◽  
Author(s):  
Richard Kwizera ◽  
Andrew Katende ◽  
Felix Bongomin ◽  
Lydia Nakiyingi ◽  
Bruce J. Kirenga
Keyword(s):  
Pulmonary Tuberculosis ◽  
Treatment Outcomes ◽  
Diagnostic Criteria ◽  
Tertiary Care ◽  
Chest Imaging ◽  
Resource Limited Settings ◽  
Pulmonary Aspergillosis ◽  
Resource Limited ◽  
Chronic Pulmonary Aspergillosis ◽  
Hiv Negative

Abstract Background Diagnosis of chronic pulmonary aspergillosis (CPA) is based on a combination of clinical symptomatology, compatible chest imaging findings, evidence of Aspergillus infection and exclusion of alternative diagnosis, all occurring for more than 3 months. Recently, a rapid, highly sensitive and specific point-of-care lateral flow device (LFD) has been introduced for the detection of Aspergillus-specific immunoglobulin (Ig)G, especially in resource-limited settings where CPA is underdiagnosed and often misdiagnosed as smear-negative pulmonary tuberculosis (PTB). Therefore, in our setting, where tuberculosis (TB) is endemic, exclusion of PTB is an important first step to the diagnosis of CPA. We used the recently published CPA diagnostic criteria for resource-limited settings to identify patients with CPA in our center. Case presentation Three Ugandan women (45/human immunodeficiency virus (HIV) negative, 53/HIV infected and 18/HIV negative), with a longstanding history of cough, chest pain, weight loss and constitutional symptoms, were clinically and radiologically diagnosed with PTB and empirically treated with an anti-tuberculous regimen despite negative microbiological tests. Repeat sputum Mycobacteria GeneXpert assays were negative for all three patients. On further evaluation, all three patients met the CPA diagnostic criteria with demonstrable thick-walled cavities and fungal balls (aspergilomas) on chest imaging and positive Aspergillus-specific IgG/IgM antibody tests. After CPA diagnosis, anti-TB drugs were safely discontinued for all patients, and they were initiated on capsules of itraconazole 200 mg twice daily with good treatment outcomes. Conclusions The availability of simple clinical diagnostic criteria for CPA and a LFD have the potential to reduce misdiagnosis of CPA and in turn improve treatment outcomes in resource-limited settings.

scholarly journals Performance of LDBio Aspergillus WB and ICT Antibody Detection in Chronic Pulmonary Aspergillosis

2021 ◽  
Vol 7 (4) ◽  
pp. 311
Author(s):  
Anna Rozaliyani ◽  
Findra Setianingrum ◽  
Sresta Azahra ◽  
Asriyani Abdullah ◽  
Ayu Eka Fatril ◽  
...  
Keyword(s):  
Receiver Operating Curve ◽  
Antibody Detection ◽  
Pulmonary Aspergillosis ◽  
Comparative Performance ◽  
Tuberculosis Patients ◽  
Chronic Pulmonary Aspergillosis ◽  
Screening Assays ◽  
Roc Area ◽  
Flow Detection ◽  
Sensitivity Specificity

The detection of Aspergillus antibody has a key role in the diagnosis of chronic pulmonary aspergillosis. Western blot (WB) and immunochromatography (ICT) lateral flow detection of Aspergillus antibody can be used as confirmatory and screening assays but their comparative performance in TB patients is not known. This study investigated the performance of these assays among 88 post-tuberculosis patients with suspected CPA. Sensitivity, specificity, receiver operating curve (ROC), area under-curve (AUC) and the agreement between two assays were evaluated. Both WB and ICT showed good sensitivity (80% and 85%, respectively) for detection of Aspergillus antibodies. Substantial agreement (0.716) between these assays was also obtained. The highest AUC result (0.804) was achieved with the combination of WB and ICT. The global intensity of WB correlated with the severity of symptoms in CPA group (p = 0.001). The combination of WB and ICT may increase specificity in CPA diagnosis.

Sign in / Sign up

Export Citation Format

Share Document