In 1912, DW Dorrance invented the Dorrance artificial hand. DW Dorrance invented the split hook artificial hand shortly before World War I. The technology continued to develop after WW1 when demand increased from soldiers returning with battle-inflicted injuries. It became popular with laborers after the war, who could return to work using the attachment because of its ability to grip and manipulate objects. It’s one of the few designs that have remained relatively unchanged over the past century.

Why did the technology change so slowly, or not at all? Did we not have the technology to make better prosthetics? Because it was unprofitable. In this case, the market for hand prosthetics was tiny, and the barrier to entry to create more functional hands was high. The treatment is a one-time sale; there is no reoccurring revenue model, which is the systemic problem with medical technology. It’s an economic problem. The barrier to entry is very high due to the cost of research and total addressable market (TAM) was low.

The same goes for many diseases that don’t have a massive number of patients (or customers for drug companies. The probability that traditional research methods will yield a profitable drug is low. However, with the inclusion of AI into the discovery and other aspects of medicine and healthcare, we can lower the barrier to entry and make treatment options less capital-intensive while improving accessibility to life-altering treatments.

Take the case of David Aguilar Amphoux, born with Poland Syndrome, which affected the development of his right arm. He ingeniously built his prosthetic arm using LEGO® bricks, becoming the first person to achieve this feat. His journey began at age nine, culminating in a functional prosthetic with a movable elbow joint and grabber at age 18. David, now 22, has created five different LEGO® prosthetic models and shares his knowledge through YouTube, aiming to make affordable prosthetics accessible and to inspire others with different abilities. His work has earned him recognition, including a Guinness World Record and invitations from NASA and LEGO's educational arm, and he continues to pursue his passion for innovation by studying bioengineering and developing his solidarity brand, HandSolo.

The takeaway is that when advanced technology becomes accessible to everyone, problems that profoundly impact individuals can be solved by more people, and innovation is democratized.

Where AI is making a difference in healthcare

Artificial intelligence (AI) promises to revolutionize nearly every industry, but perhaps none more so than healthcare. The integration of AI into healthcare has the potential to vastly improve patient outcomes while reducing costs across the entire system. From assisting doctors with complex diagnoses to optimizing hospital logistics, AI could provide healthcare professionals with insights and efficiency unequaled by traditional methods. Let’s look as some early results from AI in improving our health so far.

AI in Cancer Treatment

Cancer is a leading cause of morbidity and mortality worldwide. While progress has been made in the diagnosis, prognosis, and treatment of cancer patients, individualized and data-driven care remains a challenge. Artificial Intelligence (AI) and its subset, Machine Learning (ML), have emerged as vital tools in enhancing the accuracy and outcomes of healthcare, particularly in oncology. These technologies are increasingly used for risk assessment, early diagnosis, prognosis, and treatment selection, showing higher accuracy in predicting various cancers (like breast, brain, lung, liver, and prostate) than traditional clinical methods. AI and ML improve cancer care and have broader applications in improving diagnoses, prognoses, and quality of life for patients with various illnesses. The article underscores the importance of advancing AI and ML technologies and developing new programs for patient benefit while exploring their current uses, limitations, and potential in cancer prediction.

AI for Age-Related diseases.

Drug discovery, the process of finding new drugs, is a complex, time-consuming, and expensive task. Using AI, Researchers at the University of Edinburgh have identified three promising candidates for senolytic drugs. Senolytics are drugs that can slow down aging and prevent age-related diseases. They work by killing senescent cells, which are metabolically active cells that can no longer replicate and are called 'zombie cells.' While around 80 senolytics are known, only a combination of dasatinib and quercetin has been tested in humans. Finding more senolytics that can be used to treat a range of diseases would be helpful, but the drug discovery process can take ten to twenty years and cost billions of dollars.

The researchers employed an AI model to identify potential senolytic molecules. The model analyzed 4,340 molecules and presented a list of 21 top-scoring molecules with a high probability of being senolytics. If the researchers had tested these compounds in the lab, it would have taken several weeks of intensive work and cost £50,000 to purchase the compounds, not including the cost of experimental machinery and setup.

The researchers then tested these molecules on two types of cells: healthy and senescent. The results showed that three of the 21 compounds, namely periplocin, oleandrin, and ginkgetin, could eliminate senescent cells while keeping most normal cells alive.

AI for Heritable Diseases

At last spring's Third International Summit on Human Genome Editing, Victoria Gray, a Walmart cashier from Mississippi, underwent an experimental CRISPR gene editing procedure to treat her sickle cell disease. Gray described her life-changing experience.

Doctors harnessed the power of the revolutionary CRISPR-Cas9 technology to edit out the mutation causing Gray's illness in billions of her bone marrow cells. Once reinserted into her body, these corrected cells began pumping out normal hemoglobin to compensate for her defective type.

Beyond stories like Gray's, CRISPR promises treatments for conditions from cancer to blindness. This Nobel Prize-winning technology uses specific enzymes like Cas9 to target and edit DNA precisely. Recently, scientists discovered a new CRISPR system called Cas13 that instead targets RNA, expanding possibilities even further. To optimize Cas13 guide design, An NYU lab run by Neville Sanja’s lab collaborated with machine learning expert David Knowles. Together, they engineered a deep learning model named TIGER that was trained on CRISPR screening data to predict both on-target and off-target activity of RNA-targeting guides. TIGER outperformed previous models, providing the first tool for evaluating off-target impacts.

“Machine learning can leverage the massive datasets from modern genomics experiments,” explained Knowles. “Importantly, we used interpretable machine learning to understand why TIGER predicts certain guides will work well.”

Cas13-powered CRISPR opens doors from RNA editing to blocking expression of disease-causing genes to rapidly screening drug candidates. As Victoria Gray described firsthand, CRISPR technology brings hope to millions waiting for cures. Her smiling face at the summit encapsulated the thrilling potential of scientific ingenuity to profoundly better human life.

AI Will Upend the Medical Field

While many people are concerned about the extinction risk posed by AI, I’d like to think that the chances of you losing your life to AI are infinitesimal compared to the chance of your life being saved by a disease cured with the help of artificial intelligence.

Prompt of the Week: Health Insurance Comparison Prompt

Last month when I was in the throes of deciding how to choose my health insurance plan for 2024, I used ChatGPT’s Advanced Data Analysis for evaluating my coverage options. However, I have one more important coverage to evaluate, and that’s for Woodford the dog, I am not kidding. He’s double-insured and this year pet insurance costs are going up. So it inspired me to work on a prompt that will work equally well for my dog or a human. You simply would upload the Summary of Benefits

You will act as a health insurance expert and financial advisor. I will provide summaries of multiple benefits plans and you will provide a detailed analysis of them giving me a comparison that will help me decide which one is best for my needs. Evaluate each plan that I have uploaded and provide a comparison using the following criteria - What are the specific coverage details of each plan, including preventive care, specialist visits, hospitalization, and emergency services?"- "Provide a detailed breakdown of the costs associated with each plan, including premiums, deductibles, copayments, and out-of-pocket maximums?"- "What is the range and quality of the provider network for each plan? Are there sufficient options for specialists and facilities?"- How does each plan cover prescription medications? Are there differences in the formulary or tier system for drug coverage?"-Do the plans offer any wellness programs or incentives for preventive care?"Provide a detailed comparison including scenarios that take into variables like which plan is better should you have a catastrophic accident, which one is better for families with children, and which one is better for patients with chronic diseases(e.g. Type I Diabetes, multiple sclerosis, or Crohn's Disease.

You will get an output that looks something like this:

To get the most out of this prompt you can then start to ask additional questions or give additional information. For example, the premiums were not included in the original documents so I need to add them. I may have additional information like the provider directory and list of covered drugs which are provided in separate documents. You can continue to iterate and ask questions to make a final decision. Make sure to double-check the analysis, I haven’t seen an error but it’s entirely possible that ChatGPT may make a mistake.

Woodford conked out before he could complete his analysis but I hope this helps this complete yours.