ORTHOBIOLOGICS DIGEST™No. 8 October 2026 When AI Holds the Scalpel, AI Robotic Surgery and the Future of Orthobiologics

ORTHOBIOLOGICS DIGEST™

No. 8 | October 2026

When AI Holds the Scalpel and Other Questions About AI Robotic Surgery and the Future of Orthobiologics and the Future of Joint and Spine Preservation

Deborah Westergaard, M.D. | Physician Perspective | Regen Experts

This issue asks a recurring question: are we merely changing a pain signal, or are we understanding and supporting the structure that produced it?

COULD HUMANOID ROBOTS REPLACE SURGEONS?

When AI Holds the Scalpel

THE ARTICLE

Unitree Introducing: H2-X Kung Fu Legend Guardian;

Popular Science: “AI-powered robots can be tricked into slicing off fingers”

Humanoid robots are advancing at a remarkable pace. Machines that once struggled to walk can now run, balance, manipulate objects and perform increasingly complex physical tasks. Unitree’s H2-X demonstration makes it easy to imagine similar technology eventually assisting with technically demanding medical procedures.

A second story is more unsettling. Popular Science reported on Binghamton University researchers who demonstrated that a compromised language-model planner could redirect a robot’s task from cutting an apple to cutting a model human hand. No person was injured. The point of the experiment was to expose a safety problem: when software controls a physical machine, a software vulnerability can become a physical risk.

Read the original: https://www.youtube.com/watch?v=Y3zb4o47yqA

Additional source: https://www.popsci.com/technology/ai-powered-robot-slices-fingers/

DR. WESTERGAARD'S TAKE

I’ve spent decades performing image-guided procedures, and I think people sometimes underestimate how invasive interventional medicine can be. In orthobiologics, we may not make a large surgical incision, but we put very sharp instruments into bones, joints, tendons, ligaments, discs and other soft tissues. Some of what we do even carries surgical CPT codes.

When I place a needle into subchondral bone or access an intervertebral disc, I’m not simply giving someone a shot. I’m performing a technically demanding intervention that requires anatomy, imaging, precision and judgment.

If robots can eventually perform surgery, why wouldn’t they eventually perform orthobiologic procedures? In some ways, image-guided interventional work may be particularly well suited to robotic assistance: defined anatomy, image guidance and repeatable trajectories.

But precision is only part of medicine. What happens when the MRI doesn’t explain the symptoms? When the anatomy isn’t textbook? When the technically possible procedure is not the right procedure for that patient? That is where judgment becomes essential.

Once you give a robot authority, not just mechanical skill, the stakes change. People may say patients would never accept that, but look at how rapidly people have begun accepting self-driving vehicles. You are sitting in a machine while software makes real-time decisions with your life in its hands. We may be closer to accepting medical autonomy than we think.

And I’m not anti-robot. Frankly, I’m looking forward to the robot evolution. I would love an extra ranch hand, a robot that could deal with venomous snakes on the property, and useful personal-security technology. I’m excited about what robots may do for us.

I’m simply cautious about confusing technical capability with medical judgment. The ability to perform a procedure does not establish that the procedure is necessary, whether the instrument is held by a surgeon, an interventionalist or a robot.

Perhaps the most important question isn’t whether a robot can learn to operate. It’s whether it can learn when not to.

• • •

ARE YOUR HORMONES AGING YOUR JOINTS?

Midlife joint pain deserves more than one explanation

THE ARTICLE

CBS 17 Health Now: “Could menopause cause joint pain?” (Oct. 14, 2026)

A CBS 17 Health Now segment examined joint pain, stiffness and chronic pain during the menopause transition, including the possible role of changing hormone levels and what women should know before considering hormone therapy.

The subject deserves attention. Estrogen signaling intersects with bone, muscle and connective-tissue biology, and midlife hormonal change may contribute to musculoskeletal symptoms. But an association with menopause does not identify the pain generator in an individual knee, shoulder, hip or spine.

Read the original: https://raleighmovingnews.com/could-menopause-cause-joint-pain/

DR. WESTERGAARD'S TAKE

I think we have underestimated the relationship between hormones and musculoskeletal health for a long time. When a woman in her 40s or 50s develops knee pain, shoulder pain or stiffness, we appropriately look for arthritis, tendon injury and other structural problems. But we should also remember that the biological environment around those tissues is changing.

At the same time, I don’t want “menopause” to become another shortcut diagnosis. A woman can have changing hormones and a torn meniscus. She can have menopausal symptoms and a rotator cuff tear. She can have osteoarthritis that becomes more symptomatic during a hormonal transition. Those things are not mutually exclusive.

That distinction matters in orthobiologics. Before I consider PRP, bone marrow concentrate or any other procedure, I want to know what structure is actually hurting and why. History, examination, imaging and, when appropriate, dynamic ultrasound all matter.

Hormones may be part of the story. They are not automatically the whole story. Treat the patient, not the label.

• • •

WHAT IF WE TREATED KNEE ARTHRITIS THROUGH AN ARTERY?

A new route for cell delivery raises old questions about what we are actually treating

THE ARTICLE

University of Miami Miller School of Medicine: FDA Authorizes Novel Clinical Trial Exploring Artery-Delivered Stem Cells for Knee Osteoarthritis (Oct. 19, 2026)

The University of Miami Miller School of Medicine announced FDA authorization for an investigational clinical trial evaluating mesenchymal stem/stromal cells delivered through arteries supplying the knee in people with mild to moderate osteoarthritis who remain symptomatic after conventional nonsurgical treatment.

The cells are donor-derived from umbilical cord tissue, and the study is designed first to evaluate safety and technical feasibility. The approach grew partly from work in genicular artery embolization, which targets abnormal vessels associated with chronic synovial inflammation.

Read the original: https://news.med.miami.edu/artery-delivered-stem-cells-knee-osteoarthritis-trial-2/

DR. WESTERGAARD'S TAKE

This is an interesting concept, but I immediately have questions. Where do the cells actually go? An artery supplies a vascular territory, not necessarily the precise structure causing the patient’s pain. How much reaches synovium, subchondral bone or other target tissue? What happens in the microcirculation? And what is the intended mechanism: engraftment, signaling, anti-inflammatory effects, or something else?

I’m also interested in the connection to genicular artery embolization. I understand the rationale behind GAE. Osteoarthritis can involve abnormal neovascularization associated with synovial inflammation, and selected patients can experience meaningful pain relief after embolization.

But reducing pain and restoring the health of a joint are not necessarily the same thing.

Before intentionally obstructing blood vessels, I want to understand what is being altered, what purpose those vessels may still serve, whether the effect is reversible, and whether another approach could address the pathology while preserving tissue function. Selective embolization does not automatically mean surrounding tissue is destroyed, and the procedure has an emerging evidence base. Still, I do not consider reducing blood flow inherently regenerative simply because pain improves.

That is what makes the Miami concept intriguing. Instead of using the artery only as something to embolize, investigators are exploring it as a delivery route for a biologic product. That does not mean arterial cell delivery gets a free pass either. Vascular safety, cell characteristics, destination and mechanism all matter.

My question remains the same: are we improving the health and function of the joint, or are we simply changing the way the patient experiences the disease?

• • •

WITH PRP, DOES WHERE YOU PUT IT MATTER?

A rotator cuff study asks a very orthobiologic question

THE ARTICLE

Journal of Orthopaedic Surgery and Research: Clinical recovery and structural integrity after tendon-bone interface versus intra-articular PRP during arthroscopic rotator cuff repair (Oct. 12, 2026)

A 2026 retrospective cohort study compared 365 patients undergoing arthroscopic rotator cuff repair: PRP delivered at the tendon-bone interface and PRP delivered intra-articularly,. The study found small, transient differences in some early outcomes, with no significant pairwise differences at 12 months for the reported clinical outcomes.

The study does not establish a universal best injection site, and it is a surgical repair model rather than a nonsurgical injection trial. But it raises an important question: does anatomical delivery influence what a biologic can accomplish?

Read the original: https://link.springer.com/article/10.1186/s13018-026-07282-5

DR. WESTERGAARD'S TAKE

This resonates with me because in orthobiologics, where you deliver a biologic may matter as much as what is in the syringe.

Putting PRP into a joint space is not automatically the same thing as delivering it to the tendon attachment, ligament, meniscus or other structure that is actually failing. Biology happens in anatomy.

We also need to be careful not to overinterpret this paper. These patients had surgical rotator cuff repairs, and the findings cannot simply be transferred to nonsurgical injections. The differences were modest and largely transient.

But the underlying question is exactly the right one. Before deciding whether PRP “works,” we should ask what tissue we are trying to influence, whether the product was adequately characterized, and whether it was placed where the biology is actually needed.

Precision is not an accessory to orthobiologics. It is part of the treatment.

• • •

CAN WE 3D PRINT A NEW MENISCUS?

The future of orthobiologics may not fit inside a syringe

THE ARTICLE

Journal of Experimental Orthopaedics: Toward functional integration of 3D-printed meniscal scaffolds (Sept. 7, 2026); Journal of Orthopaedic Surgery and Research: finite element-guided 3D-printed gradient meniscus constructs (Sept. 8, 2026)

Researchers continue to explore 3D-printed meniscal scaffolds designed to reproduce the geometry, regional mechanical properties and biological environment of native meniscal tissue. A 2026 systematic review found encouraging preclinical evidence but emphasized that most work remains in animal models and that major translational challenges remain.

A separate proof-of-concept study used finite-element modeling to design a gradient hydrogel meniscus prototype. The investigators were appropriately cautious: the fabricated material did not reproduce the full mechanical properties of native meniscus, so modeled advantages cannot yet be interpreted as proof of clinical function.

Read the original: https://pmc.ncbi.nlm.nih.gov/articles/PMC13549368/

Additional source: https://link.springer.com/article/10.1186/s13018-026-07147-x

DR. WESTERGAARD'S TAKE

This one grabs me. For years, meniscal tissue was removed because it seemed expendable. We now understand how much the meniscus contributes to load distribution, joint stability and protection of cartilage and subchondral bone.

The idea of manufacturing a replacement that actually restores function is exciting. But printing something that looks like a meniscus is very different from creating one that behaves like a meniscus for years inside a human knee.

Can it reproduce the complex collagen architecture? Can it anchor securely? Can it tolerate millions of loading cycles? Will it integrate biologically? Most importantly, does it protect the cartilage and subchondral bone over time?

I love the ambition, but I want long-term human outcomes, not just beautiful prototypes.

The future of orthobiologics should not be limited to what we can put in a syringe. If tissue engineering eventually allows us to replace a damaged meniscus with something that truly restores its function, that could change the entire conversation about knee preservation.

• • •

BEFORE YOU DIM THE PAIN SIGNAL, FIND OUT WHY IT HURTS

A nerve stimulator can be technically successful and still be the wrong treatment

THE ARTICLE

Becker’s Spine Review: Orthopedic surgeons warn of unnecessary pain implants (Oct. 14, 2026)

Becker’s Spine Review reported concerns from Texas orthopedic surgeons Travis Doering, M.D., and Earl Kilbride, M.D., about peripheral nerve stimulation being used before the underlying musculoskeletal condition has been adequately evaluated.

One example involved a patient with severe shoulder arthritis who ultimately required shoulder replacement after a nerve stimulator had provided little benefit. The surgeons were not arguing against neuromodulation. Their point was that technical ability to implant a device is not a substitute for identifying the condition responsible for the pain.

Read the original: https://www.beckersspine.com/orthopedic/orthopedic-surgeons-warn-of-unnecessary-pain-implants/

DR. WESTERGAARD'S TAKE

This article speaks directly to the evolution of my own practice. I spent decades in interventional pain medicine, and I still believe neuromodulation can be valuable in the right patient. But the longer I practiced, the more important one question became: why does this patient hurt?

If a painful shoulder has advanced arthritis, a nerve stimulator does not restore the joint. If a knee is mechanically unstable, turning down the pain signal does not restore stability. And if there is a repairable structural problem, suppressing the signal may delay the treatment that actually addresses it.

Orthobiologics is not automatically the answer either. PRP and bone marrow concentrate are not substitutes for a diagnosis, and there are patients who need surgery.

My concern is a procedural mindset in which the available procedure becomes the diagnosis. Before we burn, cut, ablate, stimulate, inject or replace anything, we should ask what tissue is abnormal, what normal function it serves, and whether the proposed treatment addresses the underlying problem.

Knowing how to do a procedure is not the same as knowing when to do it.

• • •

PRP VS. STEROIDS FOR A HERNIATED DISC: DID WE ACTUALLY STUDY THE SAME THING?

A new meta-analysis may say as much about study design as it does about PRP

THE ARTICLE

Epidural platelet-rich plasma versus corticosteroid for lumbar disc herniation: a systematic review and meta-analysis of randomized controlled trials

A new systematic review and meta-analysis pooled six randomized trials involving 406 participants and found no consistent difference between epidural PRP and corticosteroid injections in pain or functional outcomes through six months. The authors themselves emphasized substantial heterogeneity, imprecision, limited sample sizes and sparse safety reporting; overall certainty of evidence ranged from very low to moderate.

The conclusion is not that PRP and corticosteroid are biologically identical. It is that the available randomized evidence does not establish a consistent difference under the protocols that were studied.

Read the original: https://link.springer.com/article/10.1186/s12891-026-10656-9

DR. WESTERGAARD'S TAKE

I have two problems with studies like this.

First, what exactly was the PRP? PRP is not one standardized drug. Preparation methods, platelet concentration, total platelet dose, leukocyte content, injected volume and other characteristics can vary. Some product information was reported in the underlying literature, but the protocols were heterogeneous enough that I am uncomfortable treating every preparation as though it were the same intervention.

But I have an even more fundamental question: were we certain these patients were hurting because of their herniated discs?

A disc herniation on MRI does not prove that it is the only pain generator. A patient may also have facet or other posterior-element pain, ligamentous pathology, stenosis, muscular dysfunction, sacroiliac pain or another contributor. Radicular pain, radiculopathy and axial discogenic pain are not interchangeable diagnoses.

And epidural PRP is not the same thing as treating the disc itself. An epidural injection may be directed at the inflammatory environment surrounding an irritated nerve root. That is a different biological objective from intradiscal treatment or treating other components of the functional spinal unit.

So I would be very cautious about taking a clinical conclusion from this meta-analysis beyond what it actually shows. Pool variable patients, variable PRP products and variable techniques, and the comparison becomes difficult to interpret.

Before asking whether the injection worked, I want to know: did we identify the right patient, the right pain generator, the right biologic, the right dose and the right target? A larger study does not fix an imprecise question.

ABOUT THE AUTHOR

Deborah Westergaard, M.D. is double board certified in Anesthesiology and Pain Medicine and is Medical Director of Regen Experts in Dallas and Plano, Texas. Her practice focuses on nonoperative spine and joint preservation, image-guided procedures and orthobiologics.

Orthobiologics Digest™ is physician commentary and education. It is not individualized medical advice. Treatments discussed may be investigational, may not be appropriate for every patient, and should not be interpreted as guarantees of outcome.

Questions Readers Are Asking

Will robots eventually perform surgery without a surgeon?Robotic systems already assist surgeons, but fully autonomous surgery raises very different questions about clinical judgment, patient selection, safety, accountability, and what happens when the planned procedure is no longer the right one.

Could robots perform PRP, bone marrow concentrate, or other orthobiologic procedures?Potentially. Many orthobiologic procedures use image guidance and precise instrument trajectories that could eventually lend themselves to robotic assistance. The harder question may be whether a robot can determine whether a procedure should be performed and which tissue actually needs treatment.

Can hormone changes cause joint pain in women?Hormonal changes may influence bone, muscle, connective tissue, inflammation, and pain perception. But new joint pain in midlife should not automatically be attributed to hormones. A structural injury or another musculoskeletal condition may be present at the same time.

Can cells be delivered through an artery to treat knee osteoarthritis?Researchers are investigating intra-arterial cell delivery for knee osteoarthritis. It remains investigational, and important questions include where the cells ultimately travel, vascular safety, mechanism of action, and whether the treatment improves joint health rather than simply symptoms.

Does where PRP is injected affect the result?It may. PRP placed inside a joint is not necessarily biologically equivalent to PRP delivered directly to an injured tendon, ligament, meniscus, tendon-bone interface, or other target. The diagnosis, PRP composition, dose, image guidance, and anatomical target may all influence outcomes.

Can scientists 3D print a replacement meniscus?Researchers are developing 3D-printed meniscal scaffolds designed to reproduce aspects of the structure and mechanics of a natural meniscus. The technology is promising but remains largely experimental, and long-term human function and durability still need to be established.

Is epidural PRP better than a steroid injection for a herniated disc?Current randomized evidence has not established a consistent advantage of epidural PRP over corticosteroid injections. However, existing studies vary in PRP preparation, dose, technique, and potentially patient selection, making broad conclusions difficult.

Does a herniated disc on MRI mean the disc is causing the patient's pain?No. Disc herniations can be present without symptoms, and patients may have more than one pain generator. Clinical history, physical examination, neurologic findings and imaging need to be considered together before attributing symptoms to a disc herniation.

Is reducing pain the same as treating the underlying joint or spine problem?Not necessarily. Pain relief can be valuable, but a lower pain score does not automatically mean that tissue health, stability, biomechanics, or the underlying pathology has improved.

How do you know whether an orthobiologic procedure is treating the right structure?

Start with the diagnosis. Identify the structure. Understand why it hurts. Then decide whether an orthobiologic procedure, another nonsurgical treatment, surgery, or no procedure at all makes sense. Treat the patient, not the picture.

Orthobiologics Digest™ | Published by Regen Experts | Deborah Westergaard, M.D. | October 24, 2026 | All rights reserved.