Week of February 12, 2024
In the office, I try to succinctly summarize patients' kidney function using the word boring. Attempting to communicate complex renal physiology with a simplified, binary, and non-specific measure (my degree of interest in changes to their health status) has some odd consequences. The most entertaining is I have inadvertently trained most patients to start office visits by asking, "Am I boring you today, Dr. Weinstein?"
---
Since last week, hospitalization rates have been trending down, but wastewater RNA concentrations are starting to rise. This trend typically means infections are not dropping, and hospitalization rates will rise again in the next 7-14 days.
The N.Y. Times COVID Tracker reflects only CDC-gathered hospital data. Hospitalization data are a (lagging) indicator.
https://www.nytimes.com/interactive/2023/us/covid-cases.html
Wastewater monitoring is more of a LEADING indicator.
------
COVID articles
I found a fantastic Twitter thread on the cellular mechanisms of coronavirus. The author routinely summarizes complex science into easily understandable tutorials, though I have been unable to discover @ejustin46's professional qualifications.
https://twitter.com/ejustin46/status/1756181629427409144
Related: https://twitter.com/erictopol/status/1755615309158043707
My patients are often older, chronically ill, and immunosuppressed. In general, I advocate for them to take Paxlovid when appropriate. MedPage Today recently published a thorough description of the data, barriers, myths, and logistics of obtaining antiviral medicine for COVID.
https://www.medpagetoday.com/infectiousdisease/covid19/108480
"CorDx announced on Thursday that it has submitted its application for the CorDx TyFast Flu A/B & COVID-19 Multiplex Rapid Test to the U.S. Food and Drug Administration for Emergency Use Authorization." Knowing you are infected is half the battle.
https://www.360dx.com/regulatory-news-fda-approvals/cordx-submits-flu-covid-19-rapid-test-fda-eua
Medical Trends and Technology
While exploring the "Taylor Swift is a government agent" theories and my ongoing struggles with anti-science amongst my patients, I found this April 2023 peer-reviewed behavioral science meta-analysis looking at the most effective ways of combating conspiracies. If nothing else, the framing of interventions and techniques, such as fact-based inoculation, pseudoscience education, and logic-based inoculation, is worth noting. And, in the spirit of TLDR, educating people to think critically and evaluate data (inoculation techniques) are the most successful means of combating misinformation. Nevertheless, debunking and rational arguments are modestly effective as well.
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0280902&s=03#pone-0280902-g002
and
While not the first AR/VR tool in medicine, Here is a newly released medical imaging app for Apple's Vision Pro: https://twitter.com/medivis_ar/status/1754885154152054907
I have no first-hand knowledge of Medivis, which makes various real-time intraoperative and teaching tools using technology to visualize imaging data. Nevertheless, their tech is an excellent example of the potential of A.R. tools in healthcare.
https://www.medivis.com/surgical-ar
Infographics
The various U.S. states' transportation departments must sense a widespread educational gap amongst the general public about pothole formation. I usually find infographics on Reddit/r/coolguides and search for the source articles or websites. I found no fewer than five different infographics on this topic. The U.K. construction company Tensar offers the best tutorial on the "birth of a pothole." Irrespective of your political, religious, or national orientation, we can all find a common enemy in potholes.
The most technically detailed:
https://www.tensar.co.uk/resources/articles/what-causes-potholes
Inferior and oversimplified pothole formation infographics (which, shockingly, skip the cracking step that allows water to get under the road surface!)
Virginia Department of Transportation
and
and
County Road Association of Michigan
and
Things I learned this week
Several loyal readers sent me another combo-interest story this week - in this case, combining my love of history with A.I. In the last few years, various scientists have scanned the charred papyrus scrolls from the town of Herculaneum (destroyed by the 79 A.D. eruption of Vesuvius, along with Pompeii) with three-dimensionally reconstructed X-ray imaging to "digitally unwrap" the fused layers. In the last few weeks, A.I. has deciphered the extracted Greek text. These scrolls are the only "preserved" ancient library and are, likely, the only remaining versions of some of the texts.
and
https://twitter.com/natfriedman/status/1754519304471814555?s=20
and
https://youtu.be/n6hHiaFDKFY?si=td2q1GNFz6D_-imk
Broad background on lost texts from antiquity
https://en.wikipedia.org/wiki/Lost_literary_work
I have never contemplated the sex lives of manta rays. And now I know rays can reproduce via parthenogenesis, with male manta rays or cross-species with sharks. This all feels very unnatural. There should be laws.
https://www.newsnationnow.com/us-news/southeast/stingray-pregnant-aquarium/
If you have any questions, you should talk to your parents or watch this YouTube video: https://www.youtube.com/watch?v=N1dDderY9As
Living with A.I.
I am finding an increasing series of articles detailing how A.I. impacts science, medicine, and learning.
While over-simplified, this thread is directionally correct - offering some thoughts on how to augment education and foundational knowledge with A.I. While I doubt ChatGPT is the "best free teacher," the ideas presented are reasonable steps (along with hefty fact-checking and critical thinking) at teaching yourself new information.
https://x.com/heyrobinai/status/1756262975013368021
A recent peer-reviewed Cell article highlights how A.I. and generative A.I. will impact biology and neuroscience. The article offers a more academic approach to the "augmenting learning" topic. Asking ChatGPT to summarize the article generates: "The article "Data science opportunities of large language models for neuroscience and biomedicine" discusses the potential of large language models (LLMs) in advancing neuroscience and biomedicine. It explores how LLMs can enrich neuroscience datasets with meta-information, bridge divides between siloed communities, fuse disparate information sources, and help understand cognitive concepts in the brain. The article delves into the evolution of language models, transformer architectures' emergence, and LLMs' scaling laws. It highlights the unprecedented transfer learning capabilities of LLMs and their applications in automated annotation, biological sequence analysis, and neuroscience research. The article emphasizes the role of LLMs in generating new insights and their potential as foundational models in various scientific domains."
https://www.cell.com/neuron/fulltext/S0896-6273(24)00042-4
The Advisory Board published a high-level review of ambient audio capture tools, allowing automated transcription and organization of exam room conversations into the medical record. While not perfect, it certainly seems
https://www.advisory.com/daily-briefing/2024/02/07/ambient-ai-tools
A.I. art of the week
In Botticelli's Birth of Venus style, draw a picture of a manta ray in a bikini, wearing a V.R. headset and tinfoil hat, and read a book entitled 'critical thinking and safe sex for sea creatures.'
https://drive.google.com/file/d/1vYGqsXQZGaFw0VRPZTMNHsXIw4F86CDR/view?usp=sharing
Clean hands and sharp minds,
Adam
Comments
Post a Comment