Biomedical researchers have been looking to expand the uses of DNA sequencing ever since the human genome project finished its initial mapping mission. That was in 2003.

Twenty years and many refinements later, AI is accelerating genomic medicine’s advance into the promised land of personalized, predictive and preventive care.

Firoz Ahmad, PhD, lays out the contours of the current landscape in an article published by The Economic Times’s Healthworld.com. Ahmad’s day job is running the clinical genomics operation at Unipath Specialty Laboratory, a pathology services concern in India. Here are four highlights of his thinking on AI in genomics and healthcare.

1. AI can speed analysis of data gleaned from genome sequencing. Ahmad points out that such analysis absent an assist from AI is not only slow and complex but also expensive and labor-intensive. How could it be otherwise? One person’s genome can generate 100 terabytes of data. With AI, researchers can accelerate the analytics process by 80 to 100 times, Ahmad explains. More:

‘AI-aided analysis helps clinicians get results quicker, determine what genetic disease a critically ill patient may have and start clinical management accordingly. In today’s world, AI has [already] made significant advances in genomics data interpretation, revolutionizing how we apply genetic data for precision medicine.’

2. AI can personalize care for people suffering with rare genetic diseases. Somewhere between 5,000 and 8,000 distinct rare diseases exist globally. Stricken patients may wait years until receiving a definitive diagnosis and, with it, an optimal treatment plan. Ahmad:

‘AI tools can match a patient’s self-reported symptoms with those of rare diseases and combine the patient’s clinical presentation, family history and high-end genomic sequencing report to predict the disease in a short time. AI-based phenotype-to-genotype mapping methodologies could greatly increase genetic testing diagnostic yield and the identification of individuals with previously unknown genetic diseases.’

3. AI can predict which cancers will metastasize and which will remain localized. This information can be crucial for optimizing treatment planning, Ahmad notes. More:

‘The combination of AI-based models with genomics data can aid in the identification of cancer subtypes, the discovery of new markers and therapeutic targets and a better knowledge of cancer-driving genes—all of which help to provide patients with personalized treatment.’

4. AI can monitor changes in a patient’s genomic data over time. This can help lock in efficacious treatments, guide care steps and save precious time, Ahmad underscores.

‘Days are not far away when a genetic abnormality could be discovered in the first few hours of a newborn’s life, ushering in a new era of next-generation diagnostics. The future of genomics is here, driven by the transformative potential of AI.’

Read the rest.

Buzzworthy developments of the past few days.

• The ChatGPT company is headed for a $1B payday. So reports The Information, which projects OpenAI’s revenues will top the B mark over the next 12 months. The Microsoft-cozy company is currently bringing in $80 million a month, so the forecast isn’t exactly pie in the sky. Fox Business notes that OpenAI had drawn a million or two subscribers to its $20/month subscription service as of March. And earlier this week it debuted ChatGPT Enterprise for businesses. The Information item is paywalled, but Fox Business’s free coverage is here.
   
• Meanwhile, a fifth of the Internet’s most visited sites are barring crawler bots from scraping their data. That includes bots dispatched by the ChatGPTs of the world. ChatGPT has said it doesn’t mind, but if the trend spreads much further, it could frustrate the company’s vision to train its algorithms on data from everywhere. Originality.ai reports that ChatGPT blocking rose from 9% to 12% over just a one-week period in August. To this Axios adds: “Media outfits that feel they got taken by Google over the past two decades are eyeing the rapid commercialization of AI services like OpenAI with hostility and a ‘we won’t get fooled again’ attitude.”
   
• Raising some tough questions about all of this are two Business Insider reporters. “How do you license the whole of the Internet?” they ask, more or less rhetorically. (Nobody seems to know the answer.) As Creative Commons CEO Catherine Stihler puts it, “So many issues keep cropping up: compensation, consent, credit. What does all of that look like with AI? … If we’re not careful, we’ll end up closing the [digital] commons.” It’s important reporting. Read the whole thing.
   
• HCA Healthcare and Google Cloud are working to bring generative AI to hospitals at scale. The effort builds on a collaboration over digital transformation that the two began in 2021. The expanded arrangement will see them tapping generative AI to relieve clinicians of administrative tasks so they can pay more attention to patients. Announcement here, more on Google’s guiding vision for broad-scale healthcare AI here.
   
• Inventors at MIT and the University of Galway (Ireland) have come up with an implantable gadget that purposefully combines AI with soft robotics. It’s designed to automatically morph, as needed, so as to avoid rejection by the patient’s unique anatomy or immune system. The creators see the device as a set-and-forget system for drug delivery aimed at sufferers of chronic conditions. The New York Post makes the advance sound like equal parts science and fun. (University news item here, peer-reviewed study here.)
   
• Healthcare AI that narcs on patients as well as prescribers? Yes, it’s here. As a matter of fact, one company’s product sports a name its miffed human targets might have given it—NarxCare. The company behind the platform, Bamboo Health, tells Kaiser’s KFF Health News NarxCare shows providers their patient’s scores for overdose risk on a single screen. However, Bamboo points out, the tool “should never replace decisions made by physicians.” Article here.
   
• Select vendor news straight from the sources:
   
  • ReferWell and Insightin announce joint offering to improve HRA (health reimbursement arrangement) performance
  • NextGen Healthcare and Luma Health announce strategic alliance to offer AI-enhanced solutions for patient engagement
  • Viz.ai signs license agreement with UCSF to commercialize three cardiac artificial intelligence algorithms
     
• From AIin.Healthcare’s news partners:
   
  • Health Imaging: Combining AI models leads to better breast cancer risk assessments
  • Cardiovascular Business: Siemens Healthineers unveils 'truly revolutionary' AI-enabled echo system
  • Health Imaging: MRI results affected by movement? MIT researchers have an AI-powered solution