Polyp Detection in Colonoscopy Images using Deep Learning and Bootstrap Aggregation

Download

Polyp Detection in Colonoscopy Images using Deep Learning and Bootstrap Aggregation.pdf

Date

2021-4-13

Author

Polat, Görkem
Işık Polat, Ece
Kayabay, Kerem
Temizel, Alptekin

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

446
views

233
downloads

Computer-aided polyp detection is playing an increasingly more important role in the colonoscopy procedure. Although many methods have been proposed to tackle the polyp detection problem, their out-of-distribution test results, which is an important indicator of their clinical readiness, are not demonstrated. In this study, we propose an ensemble-based polyp detection pipeline for detecting polyps in colonoscopy images. We train various models from EfficientDet family on both the EndoCV2021 and the Kvasir-SEG datasets, and evaluate their performances on these datasets both in- and out-of-distribution manner. The proposed architecture works in near real-time due to the efficiency of the EfficientDet architectures even when used in an ensemble setting.

Subject Keywords

Polyp detection, Colonoscopy, Medical image Processing, Deep Learning

URI

https://hdl.handle.net/11511/89910
http://ceur-ws.org/Vol-2886/paper9.pdf

Conference Name

IEEE International Symposium on Biomedical Imaging, Endoscopy Detection and Segmentation Workshop (EndoCV2021)

Collections

Graduate School of Informatics, Conference / Seminar

Suggestions

OpenMETU
Core

Selenium doped calcium phosphate biomimetic coating on Ti6Al4V orthopedic implant material for anti-cancer and anti-bacterial purposes Yılmaz, Bengi; Evis, Zafer; Banerjee, Sreeparna; Department of Biomedical Engineering (2014) In order to prevent recurrent osteosarcoma in patients undergone surgical resection for implant insertion and infection around implant site, which is a very common complication after surgery, it is very important to functionalize the surface. In this study, to combine the effective antioxidant and potential anti-cancer and anti-bacterial properties of selenium with the high biocompatibility and bioactivity of hydroxyapatite, selenium was added to the structure of hydroxyapatite via biomimetic method and coa...
Convolutional Neural Networks for Classification of Colorectal Cancer on Whole Slide Images Temena, Mehmet Arda; Isildak, Ulas; Acar, Ahmet (2021-03-27) Diagnosis of cancer is typically performed by medical pathologist visually inspecting hematoxylin and eosin (H&E) stained whole slide images. Although the early detection of various cancer types is crucial for a successful treatment, it is a challenging work due to inter- and intra-personal variability and it may often result in a disagreement between pathologists. Visual inspection of medical specimen is a well- established method for cancer diagnosis, but digital transformation is now inevitable fact, whe...
Reduced adhesion of macrophages on anodized titanium with select nanotube surface features Rajyalakshmi, Amancherla; Ercan, Batur; Balasubramanian, K.; Webster, Thomas J. (2011-01-01) One of the important prerequisites for a successful orthopedic implant apart from being osteoconductive is the elicitation of a favorable immune response that does not lead to the rejection of the implant by the host tissue. Anodization is one of the simplest surface modification processes used to create nanotextured and nanotubular features on metal oxides which has been shown to improve bone formation. Anodization of titanium (Ti) leads to the formation of TiO(2) nanotubes on the surface, and the presence...
Lung ultrasound and computed tomographic findings in pregnant woman with COVID-19 Kalafat, Erkan; Yaprak, E.; ÇINAR, GÜLE; VARLI, BULUT; Ozisik, S.; UZUN, ÇAĞLAR; AZAP, ALPAY; KOÇ, ASLI (2020-06-01) Imaging modalities play a crucial role in the management of suspected COVID-19 patients. Before reverse transcription polymerase chain reaction (RT-PCR) test results are positive, 60-93% of patients have positive chest computed tomographic (CT) findings consistent with COVID-19. We report a case of positive lung ultrasound findings consistent with COVID-19 in a woman with an initially negative RT-PCR result. The lung ultrasound-imaging findings were present between the negative and subsequent positive RT-PC...
Random effects’ distribution assumption on joint mixed modelling Özdemir, Celal Oğuz; Kalaylıoğlu Akyıldız, Zeynep Işıl; Department of Statistics (2018) Joint mixed model is an appealing approach in medical research where it is critical to estimate the odds of a fatal complication that occurs to a patient given the covariate profile such as a risk factor observed over time. For this kind of estimation, joint mixed model is used. In the standard Bayesian analysis of the model, the error variance and random effects’ variance-covariance matrix are apriori modeled independently with Inverse-Gamma and Inverse-Wishart distributions respectively. Recently however,...

Citation Formats

G. Polat, E. Işık Polat, K. Kayabay, and A. Temizel, “Polyp Detection in Colonoscopy Images using Deep Learning and Bootstrap Aggregation,” presented at the IEEE International Symposium on Biomedical Imaging, Endoscopy Detection and Segmentation Workshop (EndoCV2021), Nice, France, 2021, Accessed: 00, 2021. [Online]. Available: https://hdl.handle.net/11511/89910.