Early Prediction of Wound Healing Outcome Based on Chronic Wound Registry Database.

Chronic wounds cause a number of unnecessary amputations due to a delay in proper treatment. To expedite timely treatment, this paper presents an algorithm which uses a logistic regression classifier to predict whether the wound will heal or not within a specified time. The prediction is made at three time-points: one month, three months, and six months from the first visit of the patient to the healthcare facility. This prediction is made using a systematically collected chronic wound registry and is based entirely on data collected during patients' first visit. The algorithm achieves an area under the receiver operating characteristic curve (AUC) of 0.75, 0.72, and 0.71 for the prediction at the three time-points, respectively.Clinical relevance- Using the proposed prediction model, the clinicians will have an early estimate of the time taken to heal thereby providing appropriate treatments. We hope this will ensure timely treatments and reduce the number of unnecessary amputations.

Quantitative Comparison of Color Asymmetry Features for Automatic Melanoma Detection.

Asymmetry assessment is an important step towards melanoma detection. This paper compares some of the color asymmetry features proposed in the literature which have been used to automatically detect melanoma from color images. A total of nine features were evaluated based on their accuracy in predicting lesion asymmetry on a dataset of 277 images. In addition, the accuracies of these features in differentiating melanoma from benign lesions were compared. Results show that simple features based on the brightness difference between the two halves of the lesion performed the best in predicting asymmetry and subsequently melanoma.Clinical relevance- The proposed work will assist researchers in choosing better performing color asymmetry features thereby improving the accuracy of automatic melanoma detection. The resulting system will reduce the workload of clinicians by screening out obviously benign cases and referring only the suspicious cases to them.

Automatic Glaucoma Detection from Stereo Fundus Images.

This paper proposes a new method for automatic detection of glaucoma from stereo pair of fundus images. The basis for detecting glaucoma is using the optic cup-to-disc area ratio, where the surface area of the optic cup is segmented from the disparity map estimated from the stereo fundus image pair. More specifically, we first estimate the disparity map from the stereo image pair. Then, the optic disc is segmented from one of the stereo image. Based upon the location of the optic disc, we perform an active contour segmentation on the disparity map to segment the optic cup. Thereafter, we can compute the optic cup-to-disc area ratio by dividing the area (i.e. the total number of pixels) of the segmented optic cup region to that of the segmented optic disc region. Our experimental results using the available test dataset shows the efficacy of our proposed approach.

Efficient Deep Learning-based Wound-bed Segmentation For Mobile Applications.

This paper proposes a deep learning image segmentation method for the purpose of segmenting wound-bed regions from the background. Our contributions include proposing a fast and efficient convolutional neural networks (CNN)-based segmentation network that has much smaller number of parameters than U-Net (only 18.1% that of U-Net, and hence the trained model has much smaller file size as well). In addition, the training time of our proposed segmentation network (for the base model) is only about 40.2% of that needed to train a U-Net. Furthermore, our proposed base model also achieved better performance compared to that of the U-Net in terms of both pixel accuracy and intersection-over-union segmentation evaluation metrics. We also showed that because of the small footprint of our efficient CNN-based segmentation model, it could be deployed to run in real-time on portable and mobile devices such as an iPad.

Role of the Choroid in Automated Age-related Macular Degeneration Detection from Optical Coherence Tomography Images.

Automatic detection of age-related macular degeneration (AMD) from optical coherence tomography (OCT) images is often performed using the retinal layers only and choroid is excluded from the analysis. This is because symptoms of AMD manifest in the choroid only in the later stages and clinical literature is divided over the role of the choroid in detecting earlier stages of AMD. However, more recent clinical research suggests that choroid is affected at a much earlier stage. In the proposed work, we experimentally verify the effect of including the choroid in detecting AMD from OCT images at an intermediate stage. We propose a deep learning framework for AMD detection and compare its accuracies with and without including the choroid. Results suggest that including the choroid improves the AMD detection accuracy. In addition, the proposed method achieves an accuracy of 96.78% which is comparable to the state-of-the-art works.

Geometric corner extraction in retinal fundus images.

This paper presents a novel approach of finding corner features between retinal fundus images. Such images are relatively textureless and comprising uneven shades which render state-of-the-art approaches e.g., SIFT to be ineffective. Many of the detected features have low repeatability (<; 10%), especially when the viewing angle difference in the corresponding images is large. Our approach is based on the finding of blood vessels using a robust line fitting algorithm, and locating corner features based on the bends and intersections between the blood vessels. These corner features have proven to be superior to the state-of-the-art feature extraction methods (i.e. SIFT, SURF, Harris, Good Features To Track (GFTT) and FAST) with regard to repeatability and stability in our experiment. Overall in average, the approach has close to 10% more repeatable detected features than the second best in two corresponding retinal images in the experiment.