| 1. |
- Cirillo, Marco Domenico, 1989-
(författare)
-
A path along deep learning for medical image analysis : With focus on burn wounds and brain tumors
- 2021
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- The number of medical images that clinicians need to review on a daily basis has increased dramatically during the last decades. Since the number of clinicians has not increased as much, it is necessary to develop tools which can help doctors to work more efficiently. Deep learning is the last trend in the medical imaging field, as methods based on deep learning often outperform more traditional analysis methods. However, in medical imaging a general problem for deep learning is to obtain large, annotated datasets for training the deep networks.This thesis presents how deep learning can be used for two medical problems: assessment of burn wounds and brain tumors. The first papers present methods for analyzing 2D burn wound images; to estimate how large the burn wound is (through image segmentation) and to classify how deep a burn wound is (image classification). The last papers present methods for analyzing 3D magnetic resonance imaging (MRI) volumes containing brain tumors; to estimate how large the different parts of the tumor are (image segmentation). Since medical imaging datasets are often rather small, image augmentation is necessary to artificially increase the size of the dataset and, at the same time, the performance of a convolutional neural network. Traditional augmentation techniques simply apply operations such as rotation, scaling and elastic deformations to generate new similar images, but it is often not clear what type of augmentation that is best for a certain problem. Generative adversarial networks (GANs), on the other hand, can generate completely new images by learning the high dimensional data distribution of images and sampling from it (which can be seen as advanced augmentation). GANs can also be trained to generate images of type B from images of type A, which can be used for image segmentation. The conclusion of this thesis is that deep learning is a powerful technology that doctors can benefit from, to assess injuries and diseases more accurately and more quickly. In the end, this can lead to better healthcare for the patients.
|
|
| 2. |
- Cirillo, Marco Domenico, et al.
(författare)
-
Improving burn depth assessment for pediatric scalds by AI based on semantic segmentation of polarized light photography images
- 2021
-
Ingår i: Burns. - : Elsevier. - 0305-4179 .- 1879-1409. ; 47:7, s. 1586-1593
-
Tidskriftsartikel (refereegranskat)abstract
- This paper illustrates the efficacy of an artificial intelligence (AI) (a convolutional neural network, based on the U-Net), for the burn-depth assessment using semantic segmentation of polarized high-performance light camera images of burn wounds. The proposed method is evaluated for paediatric scald injuries to differentiate four burn wound depths: superficial partial-thickness (healing in 0–7 days), superficial to intermediate partial-thickness (healing in 8–13 days), intermediate to deep partial-thickness (healing in 14–20 days), deep partial-thickness (healing after 21 days) and full-thickness burns, based on observed healing time.In total 100 burn images were acquired. Seventeen images contained all 4 burn depths and were used to train the network. Leave-one-out cross-validation reports were generated and an accuracy and dice coefficient average of almost 97% was then obtained. After that, the remaining 83 burn-wound images were evaluated using the different network during the cross-validation, achieving an accuracy and dice coefficient, both on average 92%.This technique offers an interesting new automated alternative for clinical decision support to assess and localize burn-depths in 2D digital images. Further training and improvement of the underlying algorithm by e.g., more images, seems feasible and thus promising for the future.
|
|
| 3. |
- Cirillo, Marco Domenico, et al.
(författare)
-
Tensor Decomposition for Colour Image Segmentation of Burn Wounds
- 2019
-
Ingår i: Scientific Reports. - : NATURE PUBLISHING GROUP. - 2045-2322. ; 9
-
Tidskriftsartikel (refereegranskat)abstract
- Research in burns has been a continuing demand over the past few decades, and important advancements are still needed to facilitate more effective patient stabilization and reduce mortality rate. Burn wound assessment, which is an important task for surgical management, largely depends on the accuracy of burn area and burn depth estimates. Automated quantification of these burn parameters plays an essential role for reducing these estimate errors conventionally carried out by clinicians. The task for automated burn area calculation is known as image segmentation. In this paper, a new segmentation method for burn wound images is proposed. The proposed methods utilizes a method of tensor decomposition of colour images, based on which effective texture features can be extracted for classification. Experimental results showed that the proposed method outperforms other methods not only in terms of segmentation accuracy but also computational speed.
|
|
| 4. |
- Cirillo, Marco Domenico, et al.
(författare)
-
Time-Independent Prediction of Burn Depth using Deep Convolutional Neural Networks
- 2019
-
Ingår i: Journal of Burn Care & Research. - : Oxford University Press. - 1559-047X .- 1559-0488. ; 40:6, s. 857-863
-
Tidskriftsartikel (refereegranskat)abstract
- We present in this paper the application of deep convolutional neural networks, which are a state-of-the-art artificial intelligence (AI) approach in machine learning, for automated time-independent prediction of burn depth. Colour images of four types of burn depth injured in first few days, including normal skin and background, acquired by a TiVi camera were trained and tested with four pre-trained deep convolutional neural networks: VGG-16, GoogleNet, ResNet-50, and ResNet-101. In the end, the best 10-fold cross-validation results obtained from ResNet- 101 with an average, minimum, and maximum accuracy are 81.66%, 72.06% and 88.06%, respectively; and the average accuracy, sensitivity and specificity for the four different types of burn depth are 90.54%, 74.35% and 94.25%, respectively. The accuracy was compared to the clinical diagnosis obtained after the wound had healed. Hence, application of AI is very promising for prediction of burn depth and therefore can be a useful tool to help in guiding clinical decision and initial treatment of burn wounds.
|
|
| 5. |
- Cirillo, Marco Domenico, et al.
(författare)
-
Vox2Vox : 3D-GAN for brain tumour segmentation
- 2021
-
Ingår i: BRAINLESION: GLIOMA, MULTIPLE SCLEROSIS, STROKE AND TRAUMATIC BRAIN INJURIES (BRAINLES 2020), PT I. - Cham : Springer International Publishing. - 9783030720834 - 9783030720841 ; , s. 274-284
-
Konferensbidrag (refereegranskat)abstract
- Gliomas are the most common primary brain malignancies, with different degrees of aggressiveness, variable prognosis and various heterogeneous histological sub-regions, i.e., peritumoral edema, necrotic core, enhancing and non-enhancing tumour core. Although brain tumours can easily be detected using multi-modal MRI, accurate tumor segmentation is a challenging task. Hence, using the data provided by the BraTS Challenge 2020, we propose a 3D volume-to-volume Generative Adversarial Network for segmentation of brain tumours. The model, called Vox2Vox, generates realistic segmentation outputs from multi-channel 3D MR images, segmenting the whole, core and enhancing tumor with mean values of 87.20%, 81.14%, and 78.67% as dice scores and 6.44mm, 24.36 mm, and 18.95 mm for Hausdorff distance 95 percentile for the BraTS testing set after ensembling 10 Vox2Vox models obtained with a 10-fold cross-validation. The code is available at https://github.com/mdciri/Vox2Vox
|
|
| 6. |
- Cirillo, Marco Domenico, et al.
(författare)
-
What is the best data augmentation for 3D brain tumor segmentation?
- 2021
-
Ingår i: IEEE International Conference on Image Processing (ICIP). - : IEEE. - 9781665441155 - 9781665431026 ; , s. 36-40
-
Konferensbidrag (refereegranskat)abstract
- Training segmentation networks requires large annotated datasets, which in medical imaging can be hard to obtain. Despite this fact, data augmentation has in our opinion not been fully explored for brain tumor segmentation. In this project we apply different types of data augmentation (flipping, rotation, scaling, brightness adjustment, elastic deformation) when training a standard 3D U-Net, and demonstrate that augmentation significantly improves the network’s performance in many cases. Our conclusion is that brightness augmentation and elastic deformation work best, and that combinations of different augmentation techniques do not provide further improvement compared to only using one augmentation technique. Our code is available at https://github.com/mdciri/3D-augmentation-techniques
|
|
