Mesopotamian Journal of Artificial Intelligence in Healthcare

Measuring the Effectiveness of AI Tools in Clinical Research and Writing: A Case Study in Healthcare

Sani Salisu — 2024-01-14

This article investigates the capabilities and limitations of ChatGPT, a natural language processing (NLP) tool, and large language models (LLMs), developed from advanced artificial intelligence (AI). Designed to help computers understand and produce text understandable by humans, ChatGPT is particularly aimed at general scientific writing and healthcare research applications. Our methodology involved searching the Scopus database for ’type 2 diabetes’ and ’T2 diabetes’ articles from reputable journals. After eliminating duplicates, we used ChatGPT to formulate conclusions for each selected article by inputting their structured abstracts, excluding the original conclusions. Additionally, we tested ChatGPT’s response to simple misuse scenarios. Our findings show that ChatGPT can accurately grasp context and concisely summarize primary research findings. Additionally, it helps individuals who are not as experienced in mathematical analysis by providing coding guidelines for mathematical analyses in a variety of computer languages and by demystifying difficult model results. In conclusion, even if ChatGPT and other AI technologies are revolutionizing scientific publishing and healthcare, their use should be strictly controlled by authoritative laws.

Machine learning based Lung Disease Prediction Using Convolutional Neural Network Algorithm

M.Sahaya Sheela — 2024-06-01

Lung disease prediction is a critical issue in today's world. However, in the past two years, the corona virus disease 2019 (COVID-19) has a broad range and, in a limited percentage of people, a notable effect on the lungs. In the past, the fuzzy logic method has been used to classify lung disease prediction, but it has faced challenges such as difficulties in identifying segmented regions and output inaccuracies. To solve the issue Convolutional neural networks are used in machine learning to predict lung condition. The preprocessing based weighted average filter gives greater weight to the central value, making its contribution more significant than that of other values and can regulate the degree of image blurring. The process of segmenting based on region split and merge techniques involves separating one or more areas or entities in an image according to a size of m by n at one level of a threshold value. This segmented in multiple sub-regions of the same size, indicating a fundamental representational structure, from that Image classification using convolutional neural networks (CNNs) is a type of neural network specifically designed to extract distinct characteristics from segmented data. They are often used in tasks such as lung disease prediction and recognition due to their ability to identify intricate details in clustered data. The approach was evaluated using the MATLAB tool, a novel CNN with multiple image processing technique in our experiment to efficiently classify lung illnesses under typical circumstances, the average accuracy increased up to 97%. The results of this study show significant improvement in the prognosis of lung prediction in medical filed.

Digital Physicians: Unleashing Artificial Intelligence in Transforming Healthcare and Exploring the Future of Modern Approaches

Ban Salman Shukur — 2024-02-02

Growing global awareness that attention to health care is the basis for maintaining citizens' quality of life. Health institutions seek to increase interest in electronic care services and enhance patient results by integrating artificial intelligence techniques. Artificial intelligence tools are indispensable to diagnosis, treatment, and patient care. Integrating artificial intelligence techniques into the development of the electronic healthcare environment works to enhance public health and disease prevention and provide free services to all citizens. Designing electronic platforms raises health awareness in society, provides health programs and initiatives, and reaches homes, gardens, schools, and universities through applications based on artificial intelligence. The primary purpose of this article is to challenge the extent to which artificial intelligence is related to medicine and its contribution to the positive and negative effects of revolutionizing healthcare services.

Evaluating if Ghana's Health Institutions and Facilities Act 2011 (Act 829) Sufficiently Addresses Medical Negligence Risks from Integration of Artificial Intelligence Systems

George Benneh Mensah — 2024-02-10

With artificial intelligence (AI) integrated increasingly to enhance personalized diagnosis and data-driven treatment recommendations, this analysis examines the legal sufficiency of Ghana’s Health Institutions and Facilities Act 2011 (Act 829) to address medical negligence risks from reliance on AI systems in clinical settings. The CREAC framework structures evaluating gaps where existing health regulations may lack clarity for emerging issues of accountability. Explanation contextualizes the probabilistic nature of AI inferences and how general principles of medical negligence could have ambiguous application currently if erroneous AI contributions result in patient harm. Application to a hypothetical scenario assesses if adequate protections for appropriate integration exist across developers, systems, healthcare facilities, and practitioners under applicable interpretations of existing laws. Finding liability rules insufficient absent targeted AI governance, conclusions recommend amending Act 829 in key areas to codify expectations for responsible innovation and prevent ambiguity in liability.

This work carries scientific novelty as one of the first structured jurisdictional analyses internationally of healthcare AI accountability gaps through a legal lens. Practical significance lies in setting the stage for strengthening protections in Ghana through proposed statutory reforms that reduce uncertainty around this crucial area for quality care. The method and recommendations offer a model for modernizing medical negligence law and AI policy amidst ongoing digitization in healthcare worldwide.

Machine learning Helps in Quickly Diagnosis Cases of "New Corona"

Maad M. Mijwil — 2024-01-16

Machine learning is considered one of the most significant techniques that play a vital role in diagnosing the Coronavirus. It is a set of advanced algorithms capable of analysing medical data and identifying patterns and behaviours of diseases. It is used to interpret medical images, giving details of each image with high accuracy and efficiency, such as chest X-ray images. These algorithms are trained on a large set of images to recognise patterns that indicate the presence of infection with the Coronavirus (COVID-19). This article will provide a brief overview of the importance of machine learning in diagnosing COVID-19 by processing and analysing medical image data and helping physicians and healthcare workers provide distinguished and influential care for patients infected with this virus.

DARKNET-53 Convolutional Neural Network-Based Image Processing for Breast Cancer Detection

R. Rajkumar — 2024-06-15

Breast cancer is a common type of cancer in women, denoted by the uncontrolled growth of cells in breast tissue. Thus, manually detecting breast cancer is time-consuming and necessitates automated systems. Existing breast cancer screening methods often have limited efficacy and may delay detection and complicate the individual treatment planning process. However, early detection of breast cancer can be costly and impact the accuracy of diagnosis. To address this issue, we introduce a Darknet-53 Convolutional Neural Network (darknet-53CNN) approach for classifying breast cancer images and improving precision. Furthermore, we utilise the Contrast-Limited Adaptive Histogram Equalization (CLAHE) technique to pre-process breast cancer images to enhance image quality. Furthermore, we evaluate the intensity level of pixel images by feature extraction using the Haralick Grey-Level Co-Occurrence Matrix (HGLCM) technique. Finally, the DarkNet-53 CNN method improves the accuracy of detecting breast cancer and classifying images as benign or malignant. The proposed algorithm evaluates the specificity, sensitivity, accuracy and precision of predictive test results based on the classification of breast cancer images. Moreover, the accuracy of the proposed method has increased to 95.6% compared to the methods obtained from previous approaches.

Exploring Deep Learning Methods Used in the Medical Device Sector

Fredrick Kayusi — 2024-03-21

The healthcare sector is witnessing significant development in many aspects thanks to the effects of artificial intelligence or software, which has turned out to be the centre of attraction all over the world. This is evidence of a simple development in acquiring deep knowledge of the methods and areas in which they are used. Face detection, voice recognition, autonomous use, the defence industry, the security industry, and other fields may be displayed as examples that help complete tasks. This article surveys the impact of deep learning methods and practices in the medical device industry, and we also examine the distribution of multi-year data. It is divided into six categories: healthcare, big data and wearable technologies, biomedical code, image processing, diagnostics, and the Internet of Medical Things. As a result, the medical device industry has grown in recent years through deep learning techniques and the use of most research related to diagnosis and image processing.

Examining Ghana's Health Professions Regulatory Bodies Act, 2013 (Act 857) To Determine Its Adequacy in Governing the Use of Artificial Intelligence in Healthcare Delivery and Medical Negligence Issues

George Benneh Mensah — 2024-01-30

This analysis examines Ghana’s Health Professions Regulatory Bodies Act, 2013 (Act 857) to assess its fitness to govern the ascent of artificial intelligence (AI) in reshaping healthcare delivery. As advanced algorithms supplement or replace human judgments, dated laws centered on individual practitioner liability struggle to contemplate emerging negligence complexities. Act 857 lacks bespoke provisions for governing this new era beyond outdated assumptions of human-centric care models. With AI projected to transform medicine, proactive reforms appear vital to enable innovation gains while upholding accountability.

Through an IRAC legal analysis lens supplemented by case law spanning from the United States to Ghana, this paper demonstrates how judiciaries globally are elucidating risks from legal uncertainty given increasingly autonomous health technologies. Findings reveal governance gaps impeding equitable access to remedy where algorithmic activities contribute to patient harm. Calls for stringent training, validation and monitoring prerequisites before deploying higher-risk AI systems signal a reframed standard of care is warranted.

Detailed recommendations to modernize Act 857 and adjacent regulation are provided, covering practitioner codes, product safety, ongoing evaluation duties, and crucially, updated liability rules on apportioning fault between disparate enterprises enabling flawed AI. Beyond protecting patients and practitioners, enhanced governance can boost investor confidence in Ghana’s AI healthcare ecosystem. Ultimately astute reforms today can reinforce innovation gains tomorrow across a more ethical, accountable industry.

Evaluating ChatGPT performance in Arabic dialects: A comparative study showing defects in responding to Jordanian and Tunisian general health prompts

Malik Sallam — 2024-01-10

Background: The role of artificial intelligence (AI) is increasingly recognized to enhance digital health literacy. There is of particular importance with widespread availability and popularity of AI chatbots such as ChatGPT and its possible impact on health literacy. The involves the need to understand AI models’ performance across different languages, dialects, and cultural contexts. This study aimed to evaluate ChatGPT performance in response to prompting in two different Arabic dialects, namely Tunisian and Jordanian.

Methods: This descriptive study followed the METRICS checklist for the design and reporting of AI based studies in healthcare. Ten general health queries were translated into Tunisian and Jordanian dialects of Arabic by bilingual native speakers. The performance of two AI models, ChatGPT-3.5 and ChatGPT-4 in response to Tunisian, Jordanian, and English were evaluated using the CLEAR tool tailored for assessment of health information generated by AI models.

Results: ChatGPT-3.5 performance was categorized as average in Tunisian Arabic, with an overall CLEAR score of 2.83, compared to above average score of 3.40 in Jordanian Arabic. ChatGPT-4 showed a similar pattern with marginally better outcomes with a CLEAR score of 3.20 in Tunisian rated as average and above average performance in Jordanian with a CLEAR score of 3.53. The CLEAR components consistently showed superior performance in the Jordanian dialect for both models despite the lack of statistical significance. Using English content as a reference, the responses to both Tunisian and Jordanian dialects were significantly inferior (P<.001).

Conclusion: The findings highlight a critical dialectical performance gap in ChatGPT, underlining the need to enhance linguistic and cultural diversity in AI models’ development, particularly for health-related content. Collaborative efforts among AI developers, linguists, and healthcare professionals are needed to improve the performance of AI models across different languages, dialects, and cultural contexts. Future studies are recommended to broaden the scope across an extensive range of languages and dialects, which would help in achieving equitable access to health information across various communities.