Artificial intelligence is moving beyond hype into practical healthcare applications, with three distinct breakthroughs emerging simultaneously: AI models that can identify rare genetic disorders, electronic noses that detect cancer through skin emissions, and clinical tools that save doctors hours of paperwork per patient. These developments signal a shift from AI as a future promise to AI as a present-day tool improving patient care and clinical efficiency (Sources 1, 2, 3).

How Are AI Models Helping Diagnose Rare Diseases?

Rare diseases affect as many as 446 million people globally, yet they remain extraordinarily difficult to diagnose and treat due to sparse available data and chronic underfunding. Researchers are now deploying advanced machine learning tools to accelerate detection and treatment discovery for these conditions.

One breakthrough involves deep generative models trained on massive datasets. Rose Orenbuch, PhD, of Harvard, has developed a model called popEVE that has already identified 123 novel genetic variants potentially causing rare severe developmental disorders. Larger foundation models, which learn patterns from diverse datasets, may be able to generalize those signals to data-scarce rare disease research, potentially aiding both diagnosis and drug discovery.

"AI Models Could Improve Diagnosis and Care for Rare Diseases," according to research published in the Journal of Medical Internet Research, though experts note that aside from a few success stories, foundation models haven't yet proven their mettle in real-world clinical settings.

Simon Spichak, JMIR Correspondent

The promise is significant: by analyzing genetic data at scale, AI can uncover disease mechanisms that would take researchers years to identify manually. However, the technology remains in early stages, with most applications still in research rather than clinical deployment.

What Role Is Electronic Nose Technology Playing in Cancer Detection?

A pilot study published in the Journal of Analytical Chemistry has developed a quantum-dot-based electronic nose, or e-nose, that detects volatile organic compounds exhaled through the skin. The device uses cadmium sulfide nanocrystals to sense these chemical signatures and identify potential biomarkers indicating the presence or absence of malignant tumors.

The results are striking: the device achieved 100% accuracy and sensitivity in distinguishing cancer patients from healthy controls in the pilot study. Researchers also found the technology could potentially classify disease severity, offering a non-invasive screening method that could revolutionize early cancer detection.

This approach represents a fundamentally different diagnostic pathway. Rather than relying on imaging or blood tests, the e-nose captures biochemical signals the body naturally produces. If validated in larger clinical trials, such technology could enable rapid, low-cost cancer screening in primary care settings.

How Is AI Transforming the Doctor-Patient Encounter?

At Mayo Clinic, one of the world's most renowned hospital systems, AI is addressing a surprisingly mundane but time-consuming problem: sorting through massive patient medical records. Many patients seeking second or third opinions arrive with unsorted documents from multiple health systems, forcing physicians to manually organize and search through dozens or even hundreds of pages.

A new AI tool called Record Time automatically generates relevant patient summaries, organizes documents chronologically, and makes them searchable. Dr. Alexander Ryu, an internal medicine physician at Mayo Clinic who helped develop the system, said the tool can save between five and 30 minutes of preparation per visit, depending on case complexity. That time savings translates directly into more face-to-face time with patients and reduces the risk of missing critical details buried in medical files.

"We receive a huge volume of these records, tens of millions of pages every year, and we needed a way to find important information in that," said Dr. Alexander Ryu, Vice Chair of Innovation for the Mayo Clinic Department of Medicine.

Dr. Alexander Ryu, Vice Chair of Innovation, Mayo Clinic Department of Medicine

Mayo Clinic currently has around 150 AI models deployed within the hospital system, according to Dr. Matthew Callstrom, a radiologist and medical director of Mayo Clinic's generative AI program. The hospital is partnering with firms like Microsoft and Scale AI to develop these tools using its vast volume of patient records and research data.

Dr. Matthew Callstrom, a radiologist and medical director of Mayo Clinic's generative AI program

What Are the Broader Implications for Medical Education and Healthcare Systems?

As AI tools proliferate in clinical settings, medical schools face a critical challenge: preparing students for a tech-driven healthcare environment without allowing them to become overly dependent on AI. Katie Cottingham, writing in the Journal of Medical Internet Research, investigates how educators are integrating AI into medical training while protecting essential clinical reasoning skills.

The risks are real and specific. Experts warn against three failure modes:

• Never-skilling: Students who never develop critical thinking skills because they rely on AI from the start of their training.
• Mis-skilling: Students who develop incorrect mental models because they outsource complex reasoning to AI without understanding the underlying logic.
• De-skilling: Experienced professionals who lose clinical judgment by offloading difficult diagnostic problems to AI for quick answers.

"The brain needs to wrestle with information to make it stick. When people offload difficult problems to AI models for quick answers, they are not going through this process, and the information may not become second-nature," wrote Katie Cottingham in her analysis of AI in medical education.

Katie Cottingham, JMIR Correspondent

The solution, according to educators, is to use AI as a scaffold for learning rather than a replacement for clinical thinking. Students should engage with AI tools in ways that deepen understanding, not bypass it.

What Infrastructure Is Being Built to Scale AI Healthcare Innovation?

Beyond individual clinical tools, major research institutions are establishing dedicated centers to accelerate AI-driven healthcare discovery. Georgia Tech announced the creation of the Parker H. Petit Center for AI-Driven Health Innovation, made possible by a transformational commitment from technology entrepreneur and Georgia Tech alumnus Parker H. Petit.

The Petit Center will operate under Georgia Tech's Institute for Data Engineering and Science and bring together researchers from across the university to advance AI-driven approaches to human health. Its first research initiative will focus on creating virtual models of human cells, led by Jeffrey Skolnick, Regents' Professor and Mary and Maisie Gibson Chair in Computational Systems Biology.

"Georgia Tech has the expertise to redefine what is possible in healthcare through AI. By combining advanced computational methods with biological and medical insights, we can create powerful new approaches to predicting disease, identifying treatments, and improving patient outcomes," said Jeffrey Skolnick.

Jeffrey Skolnick, Regents' Professor and Mary and Maisie Gibson Chair, Georgia Tech School of Biological Sciences

By modeling disease at the cellular level, researchers can test therapeutic ideas faster, uncover links among different diseases, and focus on treatments most likely to help individual patients based on their unique biology. The work could expedite discovery of new therapies for some of the hardest-to-treat diseases, including pancreatic cancer and glioblastoma, an aggressive form of brain cancer.

The Petit Center's commitment will support advanced computing infrastructure, graduate and postdoctoral fellowships, seed research grants, and annual programs bringing together leading researchers from around the world working at the intersection of AI and health. Over time and with additional investments, the center's work will expand into cancer biomarker discovery, healthy aging, advanced cellular therapies, and AI-supported healthcare systems.

What Challenges Remain for AI in Clinical Settings?

Despite the promise, significant obstacles remain. Mayo Clinic's former Director of Research Operations, Traci Tamiko Eto, sued the hospital earlier this month, alleging retaliation for raising privacy and oversight concerns around some Mayo AI systems. The hospital stated it is "committed to the responsible development and deployment of AI, with privacy, security, transparency and compliance embedded throughout our processes".

Jason Droege, CEO of Scale AI, which partnered with Mayo Clinic to develop Record Time, emphasized that speed of adoption should not be the top priority in healthcare. "These predictions where everything is going to be fixed in a year or two, I think that's wildly ambitious," he said. "Quality of care is the bar, and then speed. In healthcare, you want to get it right, as fast as possible".

Jason Droege, CEO of Scale AI, which partnered with Mayo Clinic to develop Record Time

The convergence of these three developments, from rare disease AI models to electronic nose cancer detection to clinical record management tools, suggests healthcare is entering a new phase where AI augments rather than replaces human expertise. Success will depend on thoughtful integration, rigorous validation, and sustained focus on patient outcomes over technological novelty.