1
|
Vaidyanathan A, Guiot J, Zerka F, Belmans F, Van Peufflik I, Deprez L, Danthine D, Canivet G, Lambin P, Walsh S, Occchipinti M, Meunier P, Vos W, Lovinfosse P, Leijenaar RT. An externally validated fully automated deep learning algorithm to classify COVID-19 and other pneumonias on chest CT. ERJ Open Res 2022; 8:00579-2021. [PMID: 35509437 PMCID: PMC8958945 DOI: 10.1183/23120541.00579-2021] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [What about the content of this article? (0)] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/08/2021] [Accepted: 03/04/2022] [Indexed: 01/08/2023] Open
Abstract
Purpose In this study, we propose an artificial intelligence (AI) framework based on three-dimensional convolutional neural networks to classify computed tomography (CT) scans of patients with coronavirus disease 2019 (COVID-19), influenza/community-acquired pneumonia (CAP), and no infection, after automatic segmentation of the lungs and lung abnormalities. Methods The AI classification model is based on inflated three-dimensional Inception architecture and was trained and validated on retrospective data of CT images of 667 adult patients (no infection n=188, COVID-19 n=230, influenza/CAP n=249) and 210 adult patients (no infection n=70, COVID-19 n=70, influenza/CAP n=70), respectively. The model's performance was independently evaluated on an internal test set of 273 adult patients (no infection n=55, COVID-19 n= 94, influenza/CAP n=124) and an external validation set from a different centre (305 adult patients: COVID-19 n=169, no infection n=76, influenza/CAP n=60). Results The model showed excellent performance in the external validation set with area under the curve of 0.90, 0.92 and 0.92 for COVID-19, influenza/CAP and no infection, respectively. The selection of the input slices based on automatic segmentation of the abnormalities in the lung reduces analysis time (56 s per scan) and computational burden of the model. The Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) score of the proposed model is 47% (15 out of 32 TRIPOD items). Conclusion This AI solution provides rapid and accurate diagnosis in patients suspected of COVID-19 infection and influenza. A fully automated artificial intelligence-based network is proposed to classify CT volumes of patients affected with COVID-19 or influenza/CAP, and in the uninfectedhttps://bit.ly/3MJrVRi
Collapse
|
2
|
Vaidyanathan A, Guiot J, Zerka F, Deprez L, Danthine D, Bottari F, Canivet G, Lambin P, Walsh S, Occhipinti M, Meunier P, Vos W, Lovinfosse P, Leijenaar RT, This Work Has Received Support From The Eu/ef. Deep learning architecture for the classification of COVID-19 and others pneumonias sources on lung CT imaging. Imaging 2021. [DOI: 10.1183/13993003.congress-2021.oa1561] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/05/2022] Open
|
3
|
Guiot J, Vaidyanathan A, Deprez L, Zerka F, Danthine D, Frix AN, Thys M, Henket M, Canivet G, Mathieu S, Eftaxia E, Lambin P, Tsoutzidis N, Miraglio B, Walsh S, Moutschen M, Louis R, Meunier P, Vos W, Leijenaar RTH, Lovinfosse P. Development and Validation of an Automated Radiomic CT Signature for Detecting COVID-19. Diagnostics (Basel) 2020; 11:E41. [PMID: 33396587 PMCID: PMC7823620 DOI: 10.3390/diagnostics11010041] [Citation(s) in RCA: 21] [Impact Index Per Article: 5.3] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/12/2020] [Revised: 12/22/2020] [Accepted: 12/23/2020] [Indexed: 12/28/2022] Open
Abstract
The coronavirus disease 2019 (COVID-19) outbreak has reached pandemic status. Drastic measures of social distancing are enforced in society and healthcare systems are being pushed to and beyond their limits. To help in the fight against this threat on human health, a fully automated AI framework was developed to extract radiomics features from volumetric chest computed tomography (CT) exams. The detection model was developed on a dataset of 1381 patients (181 COVID-19 patients plus 1200 non COVID control patients). A second, independent dataset of 197 RT-PCR confirmed COVID-19 patients and 500 control patients was used to assess the performance of the model. Diagnostic performance was assessed by the area under the receiver operating characteristic curve (AUC). The model had an AUC of 0.882 (95% CI: 0.851-0.913) in the independent test dataset (641 patients). The optimal decision threshold, considering the cost of false negatives twice as high as the cost of false positives, resulted in an accuracy of 85.18%, a sensitivity of 69.52%, a specificity of 91.63%, a negative predictive value (NPV) of 94.46% and a positive predictive value (PPV) of 59.44%. Benchmarked against RT-PCR confirmed cases of COVID-19, our AI framework can accurately differentiate COVID-19 from routine clinical conditions in a fully automated fashion. Thus, providing rapid accurate diagnosis in patients suspected of COVID-19 infection, facilitating the timely implementation of isolation procedures and early intervention.
Collapse
Affiliation(s)
- Julien Guiot
- Department of Pneumology, University Hospital of Liège, 4020 Liège, Belgium; (A.-N.F.); (M.H.); (R.L.)
| | - Akshayaa Vaidyanathan
- Research and Development, Oncoradiomics SA, 4000 Liège, Belgium; (A.V.); (F.Z.); (N.T.); (B.M.); (S.W.); (W.V.); (R.T.H.L.)
- The D-Lab, Department of Precision Medicine, Maastricht University, 6229 Maastricht, The Netherlands;
| | - Louis Deprez
- Department of Radiology, University Hospital of Liège, 4020 Liège, Belgium; (L.D.); (D.D.); (E.E.); (P.M.)
| | - Fadila Zerka
- Research and Development, Oncoradiomics SA, 4000 Liège, Belgium; (A.V.); (F.Z.); (N.T.); (B.M.); (S.W.); (W.V.); (R.T.H.L.)
- The D-Lab, Department of Precision Medicine, Maastricht University, 6229 Maastricht, The Netherlands;
| | - Denis Danthine
- Department of Radiology, University Hospital of Liège, 4020 Liège, Belgium; (L.D.); (D.D.); (E.E.); (P.M.)
| | - Anne-Noëlle Frix
- Department of Pneumology, University Hospital of Liège, 4020 Liège, Belgium; (A.-N.F.); (M.H.); (R.L.)
| | - Marie Thys
- Department of Medico-Economic Information, University Hospital of Liège, 4020 Liège, Belgium;
| | - Monique Henket
- Department of Pneumology, University Hospital of Liège, 4020 Liège, Belgium; (A.-N.F.); (M.H.); (R.L.)
| | - Gregory Canivet
- Department of Computer Applications, University Hospital of Liège, 4020 Liège, Belgium; (G.C.); (S.M.)
| | - Stephane Mathieu
- Department of Computer Applications, University Hospital of Liège, 4020 Liège, Belgium; (G.C.); (S.M.)
| | - Evanthia Eftaxia
- Department of Radiology, University Hospital of Liège, 4020 Liège, Belgium; (L.D.); (D.D.); (E.E.); (P.M.)
| | - Philippe Lambin
- The D-Lab, Department of Precision Medicine, Maastricht University, 6229 Maastricht, The Netherlands;
| | - Nathan Tsoutzidis
- Research and Development, Oncoradiomics SA, 4000 Liège, Belgium; (A.V.); (F.Z.); (N.T.); (B.M.); (S.W.); (W.V.); (R.T.H.L.)
| | - Benjamin Miraglio
- Research and Development, Oncoradiomics SA, 4000 Liège, Belgium; (A.V.); (F.Z.); (N.T.); (B.M.); (S.W.); (W.V.); (R.T.H.L.)
| | - Sean Walsh
- Research and Development, Oncoradiomics SA, 4000 Liège, Belgium; (A.V.); (F.Z.); (N.T.); (B.M.); (S.W.); (W.V.); (R.T.H.L.)
| | - Michel Moutschen
- Department of Infectious Diseases, University Hospital of Liège, 4020 Liège, Belgium;
| | - Renaud Louis
- Department of Pneumology, University Hospital of Liège, 4020 Liège, Belgium; (A.-N.F.); (M.H.); (R.L.)
| | - Paul Meunier
- Department of Radiology, University Hospital of Liège, 4020 Liège, Belgium; (L.D.); (D.D.); (E.E.); (P.M.)
| | - Wim Vos
- Research and Development, Oncoradiomics SA, 4000 Liège, Belgium; (A.V.); (F.Z.); (N.T.); (B.M.); (S.W.); (W.V.); (R.T.H.L.)
| | - Ralph T. H. Leijenaar
- Research and Development, Oncoradiomics SA, 4000 Liège, Belgium; (A.V.); (F.Z.); (N.T.); (B.M.); (S.W.); (W.V.); (R.T.H.L.)
| | - Pierre Lovinfosse
- Department of Nuclear Medicine and Oncological Imaging, University Hospital of Liège, 4020 Liège, Belgium;
| |
Collapse
|
4
|
Damsin T, Canivet G, Jacquemin P, Seidel L, Gillet P, Giet D, Nikkels AF. Value of Teledermoscopy in Primary Healthcare Centers: Preliminary Results of the TELESPOT Project in Belgium. Dermatol Ther (Heidelb) 2020; 10:1405-1413. [PMID: 32946049 PMCID: PMC7649191 DOI: 10.1007/s13555-020-00445-0] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/13/2020] [Indexed: 11/27/2022] Open
Abstract
INTRODUCTION Teledermoscopy using smartphone-based applications is becoming more and more important in a setting of increasing frequency of skin cancer and difficult access to specialized care. The TELESPOT project aimed to provide rapid diagnosis and speed up patient flow between primary healthcare centers and a tertiary care center in Belgium. The aim of the present study is to describe the development of an in-house smartphone-based dermoscopy application, evaluate its real-life value in a series of primary healthcare centers, and present preliminary diagnostic data. METHODS Modified Likert scales were used to assess patient and general practitioner (GP) satisfaction rates for the system. Furthermore, a total of 105 photographic and dermoscopic images were acquired in a series of 80 patients at participating centers. RESULTS Overall, patient and GP satisfaction levels were 89% and 94%, respectively. High-priority management was recommended in 7.6% of cases (8/105: 3 basal cell carcinoma, 1 primary cutaneous B-cell lymphoma, 1 Spitz melanocytic nevus, 1 congenital nevus, 1 in situ melanoma, and 1 invasive melanoma, proven by histology). CONCLUSIONS The primary healthcare centers were highly satisfied with the TELESPOT project in terms of user-friendliness, efficacy, and reliability as well as in providing a reinforced image of first-line medicine efforts in combating skin cancer.
Collapse
Affiliation(s)
- Thomas Damsin
- Department of Dermatology, University Hospital Centre, CHU du Sart-Tilman, Liege, Belgium
| | - Gregory Canivet
- Department of Computer Applications, University Hospital Centre, CHU du Sart-Tilman, Liege, Belgium
| | - Pauline Jacquemin
- Department of Methods of Projects, University Hospital Centre, CHU du Sart-Tilman, Liege, Belgium
| | - Laurence Seidel
- Department of Biostatistics, University Hospital Centre, CHU du Sart-Tilman, Liege, Belgium
| | - Pierre Gillet
- Medical Director, University Hospital Centre, CHU du Sart-Tilman, Liege, Belgium
| | - Didier Giet
- Department of General Medicine, University Hospital Centre, CHU du Sart-Tilman, Liege, Belgium
| | - Arjen F Nikkels
- Department of Dermatology, University Hospital Centre, CHU du Sart-Tilman, Liege, Belgium.
| |
Collapse
|
5
|
Damsin T, Jacquemin P, Canivet G, Giet D, Gillet P, Nikkels AF. [TeleSPOT Project : early detection of melanoma by teledermoscopy in general practice]. Rev Med Liege 2019; 74:650-654. [PMID: 31833275] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [MESH Headings] [Subscribe] [Scholar Register] [Indexed: 06/10/2023]
Abstract
Since decades the incidence of skin cancer is clearly rising. This alarming trend also applies to melanoma. It represents the 4th most common cancer in women and 6th in men in 2015. Early recognition and treatment reduce both morbidity and mortality. Screening is the cornerstone of secondary prevention. However, access to reliable and rapid diagnosis is hampered by several factors, including accessibility to specialized medicine. One of the solutions to this problem is to collaborate with the first-line medical care through a teledermatology system. The TeleSPOT project, Teledermoscopy Smartphone-based Pigmented lesion diagnosis Online Taskforce, aims to provide a remote diagnostic aid by dermatologists to distinguish suspect pigmented skin lesions and accelerate their management.
Collapse
Affiliation(s)
- Th Damsin
- Service de Dermatologie, CHU Liège, Belgique
| | - P Jacquemin
- Service de Méthodologie de Projets, CHU Liège, Belgique
| | - G Canivet
- Service des Applications Informatiques, CHU Liège, Belgique
| | - D Giet
- Médecine générale, CHU Liège, Belgique
| | - P Gillet
- Direction Médicale, CHU Liège, Belgique
| | - A F Nikkels
- Service de Dermatologie, CHU Liège, Belgique
| |
Collapse
|