Showing posts with label ai. Show all posts
Showing posts with label ai. Show all posts

May 22, 2023

Algorithmic Bias in Healthcare AI

"Artificial intelligence (AI) and machine learning (ML) are used in healthcare to combat unsustainable spending and produce better outcomes with limited resources," says Ben Tuck in a recent article on the healthcare data blog The article stresses the importance of keeping algorithmic bias in check, and goes on to offer four steps to address it.

When machine learning occurs, particularly in neural network-based systems where it is essentially impossible to fully grasp what's happening within the "mind" of the AI, the system may rely on data that reflects cultural biases, such as racism, sexism, homophobia, ageism, and all of the other stereotyping structures that have become written across our languages, interests, parenting, habits - whether we can precisely identify them (or openly admit them) or not.

Tuck's post identifies two general causes, or types, of algorithmic bias: subgroup invalidity and label choice bias.

Subgroup Invalidity Bias

Subgroup invalidity arises where the AI isn't up to the task of modeling the behavior of certain subgroups, due to training on homogeneous populations. Tuck offers the example of a study of pulse oximeter algorithms that demonstrated bias as a result of training on non-diverse data. The study found that "Black patients had nearly three times the frequency of occult hypoxemia that was not detected by pulse oximetry as white patients." The possibility for adverse health outcomes is obvious.

Label Choice Bias

Label choice bias is harder to detect. This is the situation when the AI's process returns a proxy variable —a stand-in for the real thing when the target metric is unavailable. The use of cost data to predict the need for future healthcare resources is an example; because Black people experience discrimination that results in their receiving less of the care received by the White population. Cost metrics, as derived from mostly white consumers' episodes, is used as though it applies to everyone. An argument can be made that minorities receiving less acute care when needed may actually bias the model in exactly the opposite direction, and the existence of the argument is a strong reason to improve the way the model is built by including race very thoughtfully in the source investigations and in the model's computations.

Fixing It

To limit bias and make the models useful, is possible, Tuck says. "Organizations are taking major steps to ensure AI/ML is unbiased, fair, and explainable," pointing to a playbook developed by the Booth School of Business at the University of Chicago - a guide for healthcare organizations and policy makers on catching, quantifying, and reducing bias. Read Ben Tuck's article for steps that can be taken, and review the Algorithmic Bias Playbook for more on how to define, measure, and mitigate bias in AI/ML algorithms.


CarePrecise is a supplier of authoritative healthcare provider data and insights used across the healthcare community.

March 22, 2023

New Clinical Doc Software "Listens" to the Patient Visit

Microsoft's Nuance Communications has recently rolled out a new version of its clinical transcribing software, DAX Express, powered by OpenAI's GPT-4 technology. According to Microsoft, this will be the most advanced medical transcription software available in the market today. It will use natural language processing (NLP) to understand and accurately transcribe spoken words into textual information.

The GPT-4 technology is a conversational and ambient AI, which means it can understand the context of conversations and accurately transcribe verbal exchanges. This is intended to help medical professionals save time in the transcription process and make the tasks more efficient. The GPT-4 technology also allows for real-time analysis of conversations, which will help with improved accuracy in the transcription.

Nuance's ambient AI technology is designed to "listen" in during physician-patient visits and take notes. Incorporating GPT-4, DAX Express can swiftly generate draft clinical notes as soon as the patient visit concludes for expedited review from the physician or assistant. The software works fluidly with many popular electronic medical records software products, simplifying integration into existing systems.

Your next doctor's appointment may be documented by artificial intelligence, which suggests that patients should pay attention to the physician's notes section on their patient portal, to review for errors. This write has discovered inaccuracies in his own visit notes, even before the introduction of AI (or was it before AI, as there is no way to tell?).

December 29, 2022

Artificial Intelligence In Healthcare

The healthcare industry is on the brink of major transformation, thanks to healthcare-related advances in artificial intelligence. Healthcare organizations around the world, and governments, are beginning to integrate AI into their systems and processes. With AI, healthcare providers are able to improve medical diagnostics accuracy and automate administrative tasks, while improving patient care. In this blog post, we will explore how healthcare will change and the potential impact of AI on healthcare.

Advances in Medical Diagnostics

AI has the potential to revolutionize healthcare by greatly improving medical diagnostics accuracy. AI-powered tools are being used to help healthcare professionals diagnose diseases more quickly and accurately, as well as identify healthcare trends that may have previously gone unnoticed. Furthermore, AI technology can be used to monitor patient vitals in real time and detect early warning signs of disease.

Automation of Administrative Processes

The healthcare industry is full of administrative tasks that take up a considerable amount of time and resources, from filing paperwork to scheduling appointments and managing patient records. AI can automate these processes in ways that may not occur to human workers, to free up the humans to provide more focused care on the most complex cases. AI can also provide healthcare organizations with better insights into patient care and help healthcare professionals make more informed decisions.

Improved Patient Care

AI has the potential to drastically improve healthcare outcomes by providing healthcare professionals with improved data about patients, allowing them to take preemptive action or provide targeted healthcare services. AI can also be used to track healthcare trends and identify areas where healthcare quality measures could be improved.

Improved Clinician Workplaces and Opportunities

Physician offices can be made far more efficient with AI in the picture. HCC risk management is one area where AI can be used to find missed opportunities, and to strengthen Medicare reimbursement profiles. AI can often see what humans can't, either because some details just are not apparent, or because clinicians and admin personnel are overburdened with just getting through the day. In the constant struggle to keep patients as the top priority over paperwork, AI-driven systems from companies like Hindsait and MDOps can take on a share of the workload.


From the patient's perspective, will AI depersonalize medical services? If workflow streamlining cuts the wrong corners, who will suffer? Artificial intelligence, by its very nature, is a "black box." In many cases, advanced AI is very much like a person, in that it can be difficult or impossible to understand how its "thinking" works. AI needs to develop better "talk back" capability, so that human users can interrogate the system to correct errors - to ask how it is arriving at a given conclusion, and then to correct its "thinking," much as you would reshape a human employee's perceptions to obtain the most desirable outcomes. At present, such capabilities are not present, or are not being adequately utilized by the system's handlers in some environments. Busy practitioners haven't yet "merged" with these systems such that deliberate feedback is part of the clinical workflow. This will take time, and probably a few high profile mistakes. Progress here is a bit like the early progress of the medical profession. We're just now emerging from the blood-letting phase of AI, and we must hone strategies for better control of the new tools.

As HCOs begin to adopt AI-powered tools, healthcare processes, patient care and healthcare outcomes are set to improve significantly. AI will allow practitioners to diagnose diseases more accurately, automate administrative tasks, and gain insights into healthcare trends, enabling them to provide more informed and targeted care to their patients. At the end of the day, AI has the potential to revolutionize healthcare and significantly improve patient outcomes, while the cost of progress is bound to include some failure. All stakeholders, from patients to health systems to government, need to be informed as AI involvement increases, and become girded for the journey.