Article Sidebar
Published:
Jul 13, 2023DOI:
10.4314/ahs.v23i2.86Keywords:
Article Details
While African Health Sciences has been freely accessible online there have been questions on whether it is Open Access or not. We wish to clearly state that indeed African Health Sciences is Open Access. There are key issues regarding Open Access needing clarification for avoidance of doubt:
- 1. Henceforth, papers in African Health Sciences will be published under the CC BY (Creative Commons Attribution License) 4.0 International. See details on https://creativecomons.org/)
- 2. The copyright owners or the authors grant the 3rd party (perpetually and in advance) the right to disseminate, reproduce, or use the research papers in part or in full, format/medium as long as:
- No substantive errors are introduced in the process
- Attribution of authorship and correct citation details are given
- The referencing details are not changed.
Should the papers be reproduced in part, this must be clearly stated.
- 3. The papers will be freely and universally accessible online in an easily readable format such as XML in at least one widely recognized open access repository such as PUBMED CENTRAL.
B. ABRIDGED LICENCE AGREEMENT BETWEEN AUTHORS AND African Health Sciences
I submitted my manuscript to African Health Sciences and would like to affirm that:
1.0 I am authorized by my co-authors to enter into these arrangements.
2.0 I guarantee, on behalf of self and co-authors:
- That the paper is original, and has not been published in any other peer-reviewed journal; nor is it under consideration by other journal (s). It does not infringe existing copyright or any other person’s rights
- That we are/I am the sole author(s) of the paper and with authority to enter into this agreement. My granting rights to African Health Sciences is not in breach of any other obligation
- That the paper contains nothing unlawful, or libelous. Nor anything that would constitute a breach of contract, confidence or commitment given to secrecy, if published
- That I/we have taken care to ensure the integrity of the article.
3.0 I and all co-authors, agree that the paper, if accepted for publication, shall be licensed under the Creative Commons Attribution License 4.0. (see https://creativecommons.org/)
Main Article Content
Using artificial intelligence on dermatology conditions in Uganda: a case for diversity in training data sets for machine learning
Abstract
Background: In pursuit of applying universal non-biased Artificial Intelligence (AI) in healthcare, it is essential that data from different geographies are represented.
Objective: To assess the diagnostic performance of an AI-powered dermatological algorithm called Skin Image Search on Fitzpatrick
6 skin type (dark skin) dermatological conditions.
Methods: 123 dermatological images selected from a total of 173 images were retrospectively extracted from the electronic database of a Ugandan telehealth company, The Medical Concierge Group (TMCG) after getting their consent. Details of age, gender, and dermatological clinical diagnosis were analysed using R on R studio software to assess the diagnostic accuracy of the AI app along with disease diagnosis and body part. Predictability levels of the AI app were graded on a scale of 0 to 5, where 0- no prediction was made and 1-5 demonstrated a reduction incorrect diagnosis prediction rate of the AI.
Results: 76 (62%) of the dermatological images were from females and 47 (38%) from males. Overall diagnostic accuracy of the AI app on black dermatological conditions was low at 17% (21 out of 123 predictable images) compared to 69.9% performance on Caucasian skin type as reported from the training results. There were varying predictability levels correctness i.e., 1-8.9%, 2-2.4%, 3-2.4%, 4-1.6%, 5-1.6% with performance along individual diagnosis highest with dermatitis (80%).
Conclusion: There is need for diversity of image datasets used to train dermatology algorithms for AI applications to increase accuracy across skin types and geographies.
Keywords: Artificial intelligence; Dermatology; Fitzpatrick 6 skin type; Telehealth; Algorithms.
