-
Diffusion-based Data Augmentation for Skin Disease Classification: Impact Across Original Medical Datasets to Fully Synthetic Images
Authors:
Mohamed Akrout,
Bálint Gyepesi,
Péter Holló,
Adrienn Poór,
Blága Kincső,
Stephen Solis,
Katrina Cirone,
Jeremy Kawahara,
Dekker Slade,
Latif Abid,
Máté Kovács,
István Fazekas
Abstract:
Despite continued advancement in recent years, deep neural networks still rely on large amounts of training data to avoid overfitting. However, labeled training data for real-world applications such as healthcare is limited and difficult to access given longstanding privacy, and strict data sharing policies. By manipulating image datasets in the pixel or feature space, existing data augmentation t…
▽ More
Despite continued advancement in recent years, deep neural networks still rely on large amounts of training data to avoid overfitting. However, labeled training data for real-world applications such as healthcare is limited and difficult to access given longstanding privacy, and strict data sharing policies. By manipulating image datasets in the pixel or feature space, existing data augmentation techniques represent one of the effective ways to improve the quantity and diversity of training data. Here, we look to advance augmentation techniques by building upon the emerging success of text-to-image diffusion probabilistic models in augmenting the training samples of our macroscopic skin disease dataset. We do so by enabling fine-grained control of the image generation process via input text prompts. We demonstrate that this generative data augmentation approach successfully maintains a similar classification accuracy of the visual classifier even when trained on a fully synthetic skin disease dataset. Similar to recent applications of generative models, our study suggests that diffusion models are indeed effective in generating high-quality skin images that do not sacrifice the classifier performance, and can improve the augmentation of training datasets after curation.
△ Less
Submitted 11 January, 2023;
originally announced January 2023.
-
A 35-Year Longitudinal Analysis of Dermatology Patient Behavior across Economic & Cultural Manifestations in Tunisia, and the Impact of Digital Tools
Authors:
Mohamed Akrout,
Hayet Amdouni,
Amal Feriani,
Monia Kourda,
Latif Abid
Abstract:
The evolution of behavior of dermatology patients has seen significantly accelerated change over the past decade, driven by surging availability and adoption of digital tools and platforms. Through our longitudinal analysis of this behavior within Tunisia over a 35-year time frame, we identify behavioral patterns across economic and cultural dimensions and how digital tools have impacted those pat…
▽ More
The evolution of behavior of dermatology patients has seen significantly accelerated change over the past decade, driven by surging availability and adoption of digital tools and platforms. Through our longitudinal analysis of this behavior within Tunisia over a 35-year time frame, we identify behavioral patterns across economic and cultural dimensions and how digital tools have impacted those patterns in preceding years. Throughout this work, we highlight the witnessed effects of available digital tools as experienced by patients, and conclude by presenting a vision for how future tools can help address the issues identified across economic and cultural manifestations. Our analysis is further framed around three types of digital tools: "Dr. Google", social media, and artificial intelligence (AI) tools, and across three stages of clinical care: pre-visit, in-visit, and post-visit.
△ Less
Submitted 4 August, 2022;
originally announced August 2022.
-
Improving Skin Condition Classification with a Visual Symptom Checker Trained using Reinforcement Learning
Authors:
Mohamed Akrout,
Amir-massoud Farahmand,
Tory Jarmain,
Latif Abid
Abstract:
We present a visual symptom checker that combines a pre-trained Convolutional Neural Network (CNN) with a Reinforcement Learning (RL) agent as a Question Answering (QA) model. This method increases the classification confidence and accuracy of the visual symptom checker, and decreases the average number of questions asked to narrow down the differential diagnosis. A Deep Q-Network (DQN)-based RL a…
▽ More
We present a visual symptom checker that combines a pre-trained Convolutional Neural Network (CNN) with a Reinforcement Learning (RL) agent as a Question Answering (QA) model. This method increases the classification confidence and accuracy of the visual symptom checker, and decreases the average number of questions asked to narrow down the differential diagnosis. A Deep Q-Network (DQN)-based RL agent learns how to ask the patient about the presence of symptoms in order to maximize the probability of correctly identifying the underlying condition. The RL agent uses the visual information provided by CNN in addition to the answers to the asked questions to guide the QA system. We demonstrate that the RL-based approach increases the accuracy more than 20% compared to the CNN-only approach, which only uses the visual information to predict the condition. Moreover, the increased accuracy is up to 10% compared to the approach that uses the visual information provided by CNN along with a conventional decision tree-based QA system. We finally show that the RL-based approach not only outperforms the decision tree-based approach, but also narrows down the diagnosis faster in terms of the average number of asked questions.
△ Less
Submitted 7 August, 2019; v1 submitted 8 March, 2019;
originally announced March 2019.
