Posts related to technology, gadgets, cloud, informatics, or just about anything that is/can be/plugs into a computer. Relationship to radiology optional.
My 8-year-old has been into model rocketry for the past few years. Like a lot of good hobbies at that age, it has the right mix of building, anticipation, mild danger, and a satisfying countdown.
The only problem is that we live in Cleveland, and Cleveland winter is not exactly the most launch-friendly operating environment. Model rockets and lake-effect snow do not have a natural partnership. So for a good part of the winter, we talk about launches, plan launches, look at parts, imagine future launches, and wait.
The launch-day setup, finally outside after a winter of planning. That on the mount is an Estes minimum diameter rocket called Hi-Flier
Somewhere in that waiting period, the project shifted from “let’s launch rockets” to “what if we could measure what the rocket is actually doing?”
Generative AI was supposed to buy us time. An eight-month field study inside a ~200-person U.S. tech company suggests it can do the opposite: it intensifies work.
First, AI lowers skill barriers, so people take on tasks they previously wouldn’t. This means designers writing code, analysts drafting research, clinicians spinning up analyses. That feels empowering, but it also creates downstream “cleanup” work for others who must review, correct, and integrate AI-assisted output.
Second, AI makes work frictionless enough that it seeps into the in-between moments. Lunch breaks, late evenings, the quick “one more prompt.” The result is blurrier boundaries and less real recovery.
Third, AI encourages parallelism: multiple drafts, multiple threads, constant checking. That boosts throughput, but it also fragments attention.
The article goes on to describe a vicious cycle in which the more your colleagues do it, the more it becomes a culture, one in which you feel compelled to keep up. Using more AI.
I’ve felt a version of this personally. A few years ago, I stopped blogging regularly to make more time for kids and life outside work. With generative AI, getting a post out is genuinely easier. It’s a idea and some prompts, edits, and reviews away. But it still has to happen sometime… which, for me, often means well into the evening, in the dark, with the quiet (adorable) snoring noises of kids nearby.
Health systems frequently ask whether artificial intelligence will ultimately reduce the need for clinicians. Radiology provides one of the clearest real-world answers to date. AI changes the mechanics of work far more than it changes the need for expertise.
In imaging operations, AI is already being used to support exam prioritization, improve image reconstruction, and reduce friction in routine workflows. These contributions are meaningful, particularly in high-volume environments. But they do not displace clinical judgment, professional accountability, or responsibility for patient outcomes. In practice, the performance gains attributed to AI are inseparable from expert oversight and careful integration into clinical teams.
We use AI as a capacity and quality multiplier, not as a substitute for our training. That line is difficult to walk and gets thinner as products improve, but the thought matters. Deploying it primarily as a justification for workforce reduction or skill substitution introduces avoidable risk to patient safety and physician trust.
Success ultimately go to those investing in workflow, governance structures, and building solution that define clear roles for both humans and machines. Radiology may be an early example, but the underlying lesson extends well beyond imaging.
The U.S. FDA, Health Canada, and the UK’s MHRA have unveiled 10 guiding principles for Good Machine Learning Practice (GMLP) in developing AI/ML medical devices. These principles aim to ensure safety, efficacy, and quality in healthcare innovation. Key focuses include leveraging multi-disciplinary expertise, implementing good software and security practices, ensuring representative clinical study participants and data sets, maintaining independence between training and test data sets, and emphasizing the performance of the human-AI team. These guidelines also highlight the importance of clear user information, robust testing, and ongoing monitoring of deployed models to manage re-training risks and maintain performance.
In the rapidly evolving field of radiology, artificial intelligence (AI) is not just a tool but a collaborator, reshaping the dynamics of diagnosis and patient care. On the first order, the answer seemed clear: knowledge workers using AI outperforms those that don’t.
But the literature offers little detail on what happens after you embrace AI. Just with every tool ever existed, it really matters how you use it. As it turns out, it also matters how AI becomes part of your work.
To better understand this partnership, a group of Harvard Business School scholars published a study on business consultants who have, and who have not, opted to adopt GPT-4 in their daily work in spring 2023. There are several interesting conclusions – one of them delve into the analogy of centaurs versus cyborgs, concepts borrowed from mythology and science fiction that provide a vivid framework for the interaction between human intelligence and AI in radiology.
AI regulation in healthcare is coming. This article from FierceHealthcare summarizes the growing use of artificial intelligence (AI) in healthcare innovations and the increasing scrutiny from Senate lawmakers regarding AI regulation and healthcare payment. It highlights the concerns of bias in AI systems and the legislative scrutiny to ensure these technologies benefit patient care without discrimination. The discussion also covers lawsuits against major Medicare Advantage insurers for allegedly using AI to deny care, and the Centers for Medicare & Medicaid Services’ (CMS) guidance on AI use in healthcare decisions. Additionally, the need for transparency, accountability, and meaningful human review in AI applications in healthcare is emphasized, alongside calls for federal support to navigate AI’s integration into healthcare practices responsibly.
I urge you to read the full article which includes links to the first-hand sources and supplement with a short summary below for the busy professional.
1-Minute Summary
Federal lawmakers are actively discussing the impact of artificial intelligence (AI) in healthcare, emphasizing the need to protect patients from bias inherent in some big data systems without stifling innovation. These biases can discriminate against patients based on race, gender, sexual orientation, and disability.
To ensure the beneficial outcomes of AI while safeguarding patient rights, the Algorithmic Accountability Act was introduced. This act mandates healthcare systems to regularly verify that AI tools are being used as intended and are not perpetuating harmful biases, especially in federal programs like Medicare and Medicaid.
Major Medicare Advantage insurers, including Humana and UnitedHealthcare, are under legal scrutiny for allegedly using AI algorithms to deny care, highlighting the challenges of implementing AI in patient care decisions without exacerbating discrimination or introducing new biases.
The Centers for Medicare & Medicaid Services (CMS) issued guidelines prohibiting the use of AI or algorithms for making coverage decisions or denying care based solely on algorithmic predictions, emphasizing the necessity for decisions to be based on individual patient circumstances and reviewed by medical professionals.
Testimonies during the legislative hearings called for additional clarity on the proper use of AI in healthcare, suggesting the establishment of AI assurance labs for developing standards, and advocating for federal support to help healthcare organizations navigate the use of AI tools through investments in technical assistance, infrastructure, and training.
American College of Radiology
The emphasis on AI transparency as an answer to bias is not new. The American College of Radiology (ACR) recently kicked off its Transparent-AI initiative to advocate for openness and trust in AI. The program invites all manufacturers with FDA-cleared AI tools to participate. By offering detailed insights into an algorithm’s training, performance, and intended use, Transparent-AI not only boosts product credibility but also aids in integrating these innovations into diverse healthcare environments. Behind the scenes, the ACR has also advocated for transparency in AI with various federal agencies and lawmakers.
Recent research underscores a leap in neuroimaging accuracy for Alzheimer’s disease diagnosis, emphasizing the superior performance of AI-assisted radiologists over either AI or humans alone. This collaborative approach marries the meticulous precision of AI with the nuanced understanding of human experts, potentially setting a new standard in the detection of amyloid-related imaging abnormalities. Specifically, it demonstrated superior performance in detecting amyloid-related imaging abnormalities (ARIA), crucial for amyloid-β–directed antibody therapy. This synergy enhances diagnostic precision and underscores the potential of AI-enhanced radiological diagnostics to improve patient care significantly.
How will this synergy between AI and human intelligence redefine the future of medical diagnostics? Can this model be the blueprint for addressing other complex diseases? This breakthrough prompts us to envision a healthcare landscape where technology and human expertise converge to offer unparalleled patient care.
ESP8266 is a wifi enabled microcontroller. One of the most helpful ones because of it’s wifi ability and very low cost. This makes the ESP8266 popular in even commercial products that need wifi connectivity.
For development purposes, there are also a lot of variants for this chip. After some preliminary research, there appears to be two most helpful breakout boards for it.
NodeMCU
NodeMCU / ESP8266
NodeMCU is technically the name of the Lua-compatible firmware for ESP8266, which later added support for ESP32 (the more powerful, dual-core sibling of ESP8266). NodeMCU was created in 2014 when user Hong committed the first file of nodemcu-firmware to GitHub. but people sometimes use this term to refer to breakout boards using ESP8266 following this particular schema. It comes with additional chips that enable USB-to-serial and other “quality of life” enhancements that make development easier. The breakout board is also compatible with solderless breadboards, making prototyping much easier.
Research: Machine learning algorithm predicts wait time for outpatient imaging.
Commercial brain imaging AI receives FDA clearance
Paul Chang shares insight on the future of AI
Dreyer and Allen publish their views on the radiology AI ecosystem
CB Insights publishes market research on Google’s increasing involvement in healthcare AI.
Stay up to speed in 2 minutes. Radiology AI Briefing is a semi-regular series of blog posts featuring hand-picked news stories and summaries on machine learning and data science.
ACR makes several moves to advance the use of artificial intelligence in the future of medical imaging.
Researchers publish in high impact journal the successful use of AI to detect breast density.
ACR and MICCAI sign agreement to advance AI in medical imaging.
Geisinger declares successful implementation of algorithm to improve time-to-diagnosis of intracranial hemorrhages 20-fold.
Stay up to speed in 2 minutes. Radiology AI Briefing is a semi-regular series of blog posts featuring hand-picked news stories and summaries on machine learning and data science.
