How can AI assist with COVID-19 diagnosis?

Researchers have unveiled a revolutionary Artificial Intelligence (AI) system that can swiftly detect COVID-19 from chest X-rays with an impressive accuracy exceeding 98%. The findings of this study have been recently published in Nature Scientific Reports.

Professor Amir H Gandomi, the corresponding author from the University of Technology Sydney (UTS) Data Science Institute, emphasised the urgent need for efficient automated tools to detect COVID-19, given its substantial impact on public health and the global economy.

Traditional COVID-19 tests, such as real-time polymerase chain reaction (PCR), can be slow, costly, and occasionally prone to false negatives. Confirming a diagnosis often requires radiologists to manually examine CT scans or X-rays, a process that is time-consuming and susceptible to errors.

How does this AI technology work?  

The newly developed AI system holds potential, especially in regions with high COVID-19 levels facing a shortage of radiologists. Chest X-rays, being portable, widely accessible, and involving lower exposure to ionizing radiation than CT scans, make them a practical choice.

Typical COVID-19 symptoms overlap with those of the flu and other pneumonia types, making it challenging to distinguish between them. The AI system employs a Custom Convolutional Neural Network (Custom-CNN), a deep learning-based algorithm capable of swiftly and accurately distinguishing between COVID-19 cases, normal cases, and pneumonia in X-ray images.

The Custom-CNN model’s applications in medical imaging and diagnostics have demonstrated success in detecting diseases and interpreting complex medical images, potentially providing earlier and more accurate diagnoses. Professor Gandomi highlighted that deep learning provides an end-to-end solution, eliminating the need for manual biomarker searches

In situations where PCR or rapid antigen tests yield inconclusive or negative results, the AI system can be valuable for further radiological examination. Although radiologists remain crucial in medical diagnosis, AI technology can significantly aid them in making precise and efficient diagnoses.

A swift and accurate COVID-19 diagnosis is pivotal for ensuring patients receive appropriate treatment, including antivirals, which are most effective when administered within five days of symptom onset. Additionally, prompt diagnosis aids in patient isolation, minimising the risk of further infections and contributing to the reduction of pandemic outbreaks.

This breakthrough signifies a substantial advancement in addressing the ongoing challenges posed by the pandemic, potentially reshaping the landscape of COVID-19 diagnosis and management.

Generative AI in the healthcare landscape

Opinions on the integration of artificial intelligence (AI) in healthcare span a spectrum, with both optimistic and cautious perspectives shaping the discourse. 

On the optimistic side, many stakeholders believe that AI has the potential to significantly enhance efficiency and accuracy within the healthcare industry. AI algorithms like Custom-CNN can rapidly analyse extensive datasets and complex imagery, uncovering patterns and insights that may not be immediately apparent to human practitioners. This capability could lead to improved diagnostic accuracy and better patient outcomes.

The prospect of personalised medicine is also highlighted, as AI can analyse individual patient data to tailor treatment plans based on specific genetic and lifestyle factors.

Conversely, cautious views on AI in healthcare express concerns on various fronts. Ethical considerations, such as patient privacy, data security, and the potential misuse of sensitive medical information, are central to the reservations. The fear of job displacement within the healthcare workforce arises, with concerns that routine tasks currently performed by humans might become automated, impacting employment.

Critics also express reservations about the reliability and potential bias in AI algorithms. If these systems are trained on biased datasets, they may produce skewed results, leading to disparities in healthcare outcomes. Additionally, there are worries that increased reliance on AI might diminish the human connection in healthcare. The empathetic and emotional aspects of patient care may be challenging for AI to replicate, potentially impacting the quality of the patient-provider relationship and patient confidence in the health system and their treatments.

In summary, opinions on AI in healthcare are diverse, encompassing both enthusiasm for the technology’s potential benefits and caution regarding its ethical, social, and practical implications. 

As the healthcare industry continues to explore and implement AI solutions, what is the answer to the question “Could AI restore the care in healthcare?”. Only time will tell.

Opinions or facts expressed within the content have been sourced from various news sources. While every effort has been taken to source them accurately, the pharmacy, its owners, staff or other affiliates do not take any responsibility for errors in these sources. Patients should not rely on the facts or opinions in the content to manage their own health, and should seek the advice of an appropriate medical professional. Further, the opinions or facts in the content do not reflect the opinions and beliefs of the pharmacy, its owners, staff or other affiliates. 

Leave a Reply

Your email address will not be published. Required fields are marked *