Medicine × LLM × R

Keynote · R Conference Malaysia 2026, Sunway University

Dr. Wan Nor Arifin

Department of Biostatistics & Research Methodology, Universiti Sains Malaysia

wnarifin.github.io

21 November 2026

Outlines

  • Medicine and computers: from computer-aided diagnosis to AI

  • The rise of large language models (LLMs) and what they do for medicine

  • Where stands: packages and showcases

  • Take-home messages

Medicine × LLM ×

Part 1
Medicine and computers: from computer-aided diagnosis to AI

Medical tasks that call for computer assistance

Diagnosis

Differential diagnosis, probabilistic reasoning (Ledley and Lusted 1959; de Dombal et al. 1972)

Medical imaging

Detection and classification in radiology, retina, skin (Doi 2007; Gulshan et al. 2016)

Decision support

Alerts, order checking, guideline reminders (Bates et al. 1998; Kawamoto et al. 2005)

Prediction

Risk scores and early warning, e.g. sepsis (Wong et al. 2021)

Documentation

Records, notes, discharge summaries: a major source of clinician burden

Evidence synthesis

Searching, screening and extracting data from studies (Gartlehner et al. 2024)

Discovery

Protein structure, drug targets (Jumper et al. 2021)

Research & education

Statistical analysis, teaching, assessment

Common thread: large volume, complex reasoning, repetitive work, where humans tire and err.

From computer-aided diagnosis to AI: a timeline

Ledley and Lusted (1959); de Dombal et al. (1972); Shortliffe (1976); Miller, Pople, and Myers (1982); Bates et al. (1998); Doi (2007); Kawamoto et al. (2005); Gulshan et al. (2016); Abràmoff et al. (2018); McKinney et al. (2020); Wynants et al. (2020); Jumper et al. (2021)

Era 1: rules and probabilities (1959–1980s)

  • Ledley and Lusted (1959): diagnosis as symbolic logic + probability (Bayes) + value theory
  • de Dombal et al. (1972): prospective real-time comparison in 304 patients with acute abdominal pain
    • Computer accuracy 91.8% vs most senior clinician 79.6%
  • MYCIN (Shortliffe 1976): if–then rules for infections and antibiotic choice
  • INTERNIST-1 (Miller, Pople, and Myers 1982): diagnostic consultant for general internal medicine

Bayes’ theorem was the first “AI” at the bedside

Knowledge was hand-coded by experts (expert system): powerful but brittle and hard to maintain

Era 2: digitisation and decision support (1990s–2000s)

Clinical decision support

  • Computerised physician order entry: nonintercepted serious medication errors ↓ 55% (Bates et al. 1998)
  • Systematic review of 70 RCTs: decision support improved practice in 68% of trials; best when automatic and within workflow (Kawamoto et al. 2005)

CAD in imaging

  • Mammography CAD approved by FDA in 1998 (Doi 2007)
  • But in real-world use, sensitivity ↑ (not sig.), specificity ↓, and biopsy rate ↑ 19.7%; increased potential harms of screening mammography with uncertain clinical benefit (Fenton et al. 2007)

Lesson: workflow fit and real-world evaluation decide success

Era 3: deep learning (2012–today)

A cautionary tale

A widely implemented proprietary sepsis model: AUC 0.63 on external validation, missed 67% of sepsis cases, but alerts for 18% of all inpatients – alert fatigue (Wong et al. 2021)

Reviews: Topol (2019); Rajpurkar et al. (2022)

2020: the COVID-19 AI boom … and its lessons

The promise

  • Urgent need: RT-PCR took up to 2 days and test kits were short (Mei et al. 2020)
  • AI on chest CT and X-ray looked like a fast answer
  • Within months: hundreds of diagnostic and prognostic models

The reality

  • Systematic review of 731 prediction models: “Most published prediction model studies were poorly reported and at high risk of bias such that their reported predictive performances are probably optimistic” (Wynants et al. 2020)
  • Systematic review of 62 studies on machine learning for COVID-19 imaging: none “… are of potential clinical use due to methodological flaws and/or underlying biases” (Roberts et al. 2021)

Speed without methodology gives models that cannot be used

Lessons from 60 years of computers in medicine

  1. Computers are assistants: they extend, not replace, clinicians
  2. Workflow integration matters as much as accuracy
  3. Evaluation in real-world settings is non-negotiable: sensitivity, specificity, calibration, impact
  4. Most of the remaining burden is text: notes, letters, reports, papers

… and text is exactly what large language models are good at

Part 2
The rise of large language models

From “Attention is all you need” to today

Vaswani et al. (2017); Radford et al. (2018); Devlin et al. (2019); Radford et al. (2019); Brown et al. (2020); Ouyang et al. (2022); OpenAI (2022); OpenAI (2023); Touvron et al. (2023); Singhal et al. (2023); Singhal et al. (2025); Guo et al. (2025); Sellergren et al. (2025); Google Research and Google DeepMind (2026); McDuff et al. (2025); O’Sullivan et al. (2026)

Early signals: LLMs know (a lot of) medicine

Medical exams

Communication with patients

  • 195 patient questions from a public forum: evaluators preferred chatbot responses to physician responses in 78.6% evaluations – better quality, more empathic than human physicians (Ayers et al. 2023)

But …

Of 519 LLM evaluation studies (2022–2024), only 5% used real patient-care data (Bedi et al. 2025). So, is it valid?

How medicine benefits from LLMs: by task

Task Example study Key finding
Diagnosis AMIE, 302 differential diagnoses of complex cases (McDuff et al. 2025) Top-10 accuracy: AMIE alone 59.1% vs unassisted clinicians 33.6%
Physician RCT (Goh et al. 2024) LLM access did not improve physicians’ diagnostic reasoning; but the LLM alone scored 16 percentage points higher than physicians with conventional resources
Cardiology RCT (O’Sullivan et al. 2026) Fewer clinical errors with AMIE assistance (13.1% vs 24.3%)
Documentation Ambient AI scribes, RCTs (Lukac et al. 2025; Afshar et al. 2025) Less time in notes, less work exhaustion; occasional inaccuracies
Summarisation 4 clinical summarisation tasks (Van Veen et al. 2024) LLM summaries were superior to medical experts’ in 36% of cases (equivalent in 45%)
Information extraction 1026 lung cancer, 191 pediatric osteosarcoma pathology reports (Huang et al. 2024) classification accuracy: 89% and 98.6% respectively
Evidence synthesis Data extraction from RCTs for systematic review (Gartlehner et al. 2024) 96.3% accuracy over 160 data elements from 10 papers
Dialogue / history-taking Conversational AMIE (Tu et al. 2025) Higher diagnostic accuracy than primary care physicians in text-based consultations

Human + AI collaboration

  • RCT of 70 clinicians: AI as first or second opinion improved diagnostic accuracy (85% and 82% vs 75% with conventional resources) – complementary role of AI (Everett et al. 2026)
  • Just giving clinicians a chatbot was not enough – emphasis on AI as assistant (Goh et al. 2024)

An LLM is an assistant, not a replacement

Risks we must manage

  • Privacy: patient data sent to cloud providers
  • Changing models: behaviour shifts between versions, opaque training data (Iozzia 2026)
  • Evaluation gap: few studies on real patient data (Bedi et al. 2025)

These are statistical and governance problems, not only engineering ones

Part 3
Where stands

Python dominates LLM development

  • Pre-training, fine-tuning and serving: PyTorch, Hugging Face Transformers, vLLM, …
  • In 2024, Python overtook JavaScript as the most used language on GitHub, driven by AI and data science (GitHub 2024)
  • Practitioner LLM books assume Python and PyTorch skills, e.g. Iozzia (2026)

So… is still relevant for LLMs in medicine?

Most medical users do not build LLMs; they use them

Build the engine

  • Pre-training from scratch
  • Fine-tuning, quantisation
  • Serving at scale

Drive the car

  • Inference: prompt a model via API or locally
  • Structure text into analysable data
  • Evaluate the outputs statistically
  • Report and share: Quarto, Shiny

For the second column, is a first-class citizen

Where stands

1. Inference

Cloud APIs (OpenAI, Claude, Gemini, …) and local models via Ollama: same R session, one interface (Wickham et al. 2026; Gruber and Weber 2025)

2. Prototyping research ideas

Test an idea in minutes, next to your data, your models and your plots

3. Text → data

Extract structured variables from notes and papers straight into data frames (Brüll 2026; Arifin 2026b)

4. Evaluation

Accuracy, agreement, CIs, trials of LLM tools: biostatistics’ home turf (Couch 2025d)

5. Simple apps

Shiny + chat UI for patient education or data Q&A in a few lines (Cheng, Sievert, and Aden-Buie 2026)

6. Reproducible reporting

Prompts, model versions, seeds and results in one Quarto document

7. Privacy-first local LLMs

Patient data stay on your machine with Ollama + R (Arifin 2026a)

8. Trusted in clinical research

R-based submission packages reviewed by the FDA (R Consortium 2024)

9. RAG, tools & agents

Retrieval over guidelines, tool calling, MCP: R functions as tools (Kalinowski and Falbel 2026; Couch 2025c)

The + LLM package landscape

Purpose Packages
Unified chat interface (cloud + local) ellmer (Wickham et al. 2026), tidyllm (Brüll 2026)
Local models via Ollama rollama (Gruber and Weber 2025), ollamar (Lin and Safi 2025)
LLM over data frame columns mall (Ruiz 2025b)
Retrieval-augmented generation ragnar (Kalinowski and Falbel 2026), rchroma
Evaluation vitals (Couch 2025d)
Apps shinychat (Cheng, Sievert, and Aden-Buie 2026), querychat (natural language → SQL over data)
Coding assistants btw (Aden-Buie, Couch, and Cheng 2026), gander (Couch 2025b), chores (Couch 2025a), chattr (Ruiz 2025a), gptstudio (Nivard, Wade, and Calderon 2025)
Model Context Protocol mcptools (Couch 2025c): let assistants use your R session, and R use MCP tools
Vision kuzco (computer vision with LLMs) (Hull 2025)

Today’s showcase: rollama → tidyllm → ellmer, from basic to more advanced

Showcase: tasks from Parts 1 & 2, done in

# Medical task (Parts 1 & 2) Package What we do
1 Patient communication, translation rollama Explain a finding in simple Malay, locally
2 Triage / text classification rollama Zero-shot classification of complaints
3 Information extraction tidyllm Structure clinical notes into a data frame
4 Extraction + evaluation ellmer Extract labs from many notes, check against gold standard
5 Decision support (calculation) ellmer Tool calling: R computes eGFR, the LLM explains
6 Patient education, grounded in a document ragnar Retrieval (RAG) over a Malay Wikipedia page
7 Simple app ellmer + shinychat + ragnar Patient-education chatbot that cites the page

All examples use open-weight models through Ollama on a local PC, except where noted. Full code and output: demo document (next slides).

Showcase

rollama · tidyllm · ellmer · ragnar · shinychat

Showcase: demo document

  1. rollama: basic local inference
  2. rollama: zero-shot triage classification
  3. tidyllm: structuring clinical notes
  4. ellmer: extraction at scale, then evaluation
  5. ellmer: tool calling (eGFR in )
  6. ragnar: grounding answers in a document (RAG)
  7. ellmer + shinychat + ragnar: a patient-education app

Practical considerations

Privacy & ethics

  • Identifiable patient data → local models or approved, contracted services only
  • Ethics approval and data governance as for any data processing

Reproducibility

  • Record model name + version, prompts, seed, temperature
  • Keep prompts and code in Quarto / Git

Validation

  • Gold standard on a sample; report accuracy with CIs, agreement, errors by type
  • Re-validate when the model changes

Resources

  • Local: gaming-grade GPU for 3–14B models
  • Cloud: cost per token, rate limits, downtime

Learn more

Exploring Large Language Models With

wnarifin.github.io/explore_llm_with_r (Arifin 2026a)

  • Local LLMs, text generation, vision
  • Embeddings, classification
  • RAG, tool calling and agents
  • Fine-tuning and “LLM from scratch” (in progress)

Workshop materials

Structuring clinical notes with and LLMs: wnarifin.github.io/workshop.html (Arifin 2026b)

Take-home messages

  1. Computers have assisted medicine for 60+ years; LLMs are the newest assistant, and they speak text
  2. LLMs already help with diagnosis, documentation, summarisation and extraction, but evaluation on real data lags behind
  3. Python builds the models; is highly relevant for using them in medicine: inference, prototyping, text → data, evaluation, apps, reporting
  4. Start with rollama (local), structure data with tidyllm, grow with ellmer and ragnar
  5. Our biostatistics strengths, validation, reproducibility, inference, are exactly what medical LLM work needs

Medicine × LLM × : the centre of the Venn diagram is ours to build

References

Abràmoff, Michael D, Philip T Lavin, Michele Birch, Nilay Shah, and James C Folk. 2018. “Pivotal trial of an autonomous AI-based diagnostic system for detection of diabetic retinopathy in primary care offices.” NPJ Digit Med 1: 39. https://doi.org/10.1038/s41746-018-0040-6.
Aden-Buie, Garrick, Simon Couch, and Joe Cheng. 2026. Btw: A Toolkit for Connecting R and Large Language Models. https://posit-dev.github.io/btw/.
Afshar, Majid, Mary Ryan Baumann, Felice Resnik, Josie Hintzke, Anne Gravel Sullivan, et al. 2025. “A Pragmatic Randomized Controlled Trial of Ambient Artificial Intelligence to Improve Health Practitioner Well-Being.” NEJM AI 2 (12). https://doi.org/10.1056/aioa2500945.
Arifin, Wan Nor. 2026a. “Exploring Large Language Models with R.” https://wnarifin.github.io/explore_llm_with_r/.
———. 2026b. “Structuring Clinical Notes into Analyzable Data with R and LLMs.” https://wnarifin.github.io/workshop.html.
Ayers, John W, Adam Poliak, Mark Dredze, Eric C Leas, Zechariah Zhu, Jessica B Kelley, Dennis J Faix, et al. 2023. “Comparing Physician and Artificial Intelligence Chatbot Responses to Patient Questions Posted to a Public Social Media Forum.” JAMA Intern Med 183 (6): 589–96. https://doi.org/10.1001/jamainternmed.2023.1838.
Bates, D W, L L Leape, D J Cullen, N Laird, L A Petersen, J M Teich, E Burdick, et al. 1998. “Effect of computerized physician order entry and a team intervention on prevention of serious medication errors.” JAMA 280 (15): 1311–16. https://doi.org/10.1001/jama.280.15.1311.
Bedi, Suhana, Yutong Liu, Lucy Orr-Ewing, Dev Dash, Sanmi Koyejo, Alison Callahan, Jason A Fries, et al. 2025. “Testing and Evaluation of Health Care Applications of Large Language Models: A Systematic Review.” JAMA 333 (4): 319–28. https://doi.org/10.1001/jama.2024.21700.
Brown, Tom B., Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared Kaplan, Prafulla Dhariwal, et al. 2020. “Language Models Are Few-Shot Learners.” In Advances in Neural Information Processing Systems 33 (NeurIPS 2020). https://arxiv.org/abs/2005.14165.
Brüll, Eduard. 2026. Tidyllm: Tidy Integration of Large Language Models. https://edubruell.github.io/tidyllm/.
Cheng, Joe, Carson Sievert, and Garrick Aden-Buie. 2026. Shinychat: Chat UI Component for Shiny. https://posit-dev.github.io/shinychat/.
Couch, Simon. 2025a. Chores: A Collection of Large Language Model Assistants. https://simonpcouch.github.io/chores/.
———. 2025b. Gander: High Performance, Low Friction Large Language Model Chat. https://simonpcouch.github.io/gander/.
———. 2025c. Mcptools: Model Context Protocol for R. https://posit-dev.github.io/mcptools/.
———. 2025d. Vitals: Large Language Model Evaluation. https://vitals.tidyverse.org.
de Dombal, F T, D J Leaper, J R Staniland, A P McCann, and J C Horrocks. 1972. “Computer-aided diagnosis of acute abdominal pain.” Br Med J 2 (5804): 9–13. https://doi.org/10.1136/bmj.2.5804.9.
Devlin, Jacob, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. “BERT: Pre-Training of Deep Bidirectional Transformers for Language Understanding.” In Proceedings of NAACL-HLT 2019, 4171–86. https://arxiv.org/abs/1810.04805.
Doi, Kunio. 2007. “Computer-aided diagnosis in medical imaging: historical review, current status and future potential.” Comput Med Imaging Graph 31 (4-5): 198–211. https://doi.org/10.1016/j.compmedimag.2007.02.002.
Esteva, Andre, Brett Kuprel, Roberto A Novoa, Justin Ko, Susan M Swetter, Helen M Blau, and Sebastian Thrun. 2017. “Dermatologist-level classification of skin cancer with deep neural networks.” Nature 542 (7639): 115–18. https://doi.org/10.1038/nature21056.
Everett, Selin S, Bryan J Bunning, Priyank Jain, Ivan Lopez, et al. 2026. “From tool to teammate in a randomized controlled trial of clinician-AI collaborative workflows for diagnosis.” NPJ Digit Med 9 (1). https://doi.org/10.1038/s41746-026-02545-1.
Fenton, Joshua J, Stephen H Taplin, Patricia A Carney, Linn Abraham, Edward A Sickles, Carl D’Orsi, Eric A Berns, et al. 2007. “Influence of computer-aided detection on performance of screening mammography.” N Engl J Med 356 (14): 1399–409. https://doi.org/10.1056/NEJMoa066099.
Gartlehner, Gerald, Leila Kahwati, Rainer Hilscher, Ian Thomas, Shannon Kugley, Karen Crotty, Meera Viswanathan, et al. 2024. “Data extraction for evidence synthesis using a large language model: A proof-of-concept study.” Res Synth Methods 15 (4): 576–89. https://doi.org/10.1002/jrsm.1710.
GitHub. 2024. “Octoverse: AI Leads Python to Top Language as the Number of Global Developers Surges.” https://github.blog/news-insights/octoverse/octoverse-2024/.
Goh, Ethan, Robert Gallo, Jason Hom, Eric Strong, Yingjie Weng, Hannah Kerman, Joséphine A Cool, et al. 2024. “Large Language Model Influence on Diagnostic Reasoning: A Randomized Clinical Trial.” JAMA Netw Open 7 (10): e2440969. https://doi.org/10.1001/jamanetworkopen.2024.40969.
Google Research and Google DeepMind. 2026. “MedGemma 1.5 Technical Report.” https://arxiv.org/abs/2604.05081.
Gruber, Johannes B., and Maximilian Weber. 2025. Rollama: Communicate with Ollama to Run Large Language Models Locally. https://jbgruber.github.io/rollama/.
Gulshan, Varun, Lily Peng, Marc Coram, Martin C Stumpe, Derek Wu, Arunachalam Narayanaswamy, Subhashini Venugopalan, et al. 2016. “Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs.” JAMA 316 (22): 2402–10. https://doi.org/10.1001/jama.2016.17216.
Guo, Daya, Dejian Yang, Haowei Zhang, Junxiao Song, Peiyi Wang, Qihao Zhu, Runxin Xu, et al. 2025. “DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning.” Nature 645 (8081): 633–38. https://doi.org/10.1038/s41586-025-09422-z.
Huang, Jingwei, Donghan M Yang, Ruichen Rong, Kuroush Nezafati, Colin Treager, et al. 2024. “A critical assessment of using ChatGPT for extracting structured data from clinical notes.” NPJ Digit Med 7 (1): 106. https://doi.org/10.1038/s41746-024-01079-8.
Hull, Frank. 2025. Kuzco: Computer Vision with Large Language Models. https://frankiethull.github.io/kuzco/.
Iozzia, Guglielmo. 2026. “Domain-Specific Small Language Models: Efficient AI for Local Deployment.” Manning.
Jumper, John, Richard Evans, Alexander Pritzel, Tim Green, Michael Figurnov, Olaf Ronneberger, Kathryn Tunyasuvunakool, et al. 2021. “Highly accurate protein structure prediction with AlphaFold.” Nature 596 (7873): 583–89. https://doi.org/10.1038/s41586-021-03819-2.
Kalinowski, Tomasz, and Daniel Falbel. 2026. Ragnar: Retrieval-Augmented Generation (RAG) Workflows. https://ragnar.tidyverse.org.
Kawamoto, Kensaku, Caitlin A Houlihan, E Andrew Balas, and David F Lobach. 2005. “Improving clinical practice using clinical decision support systems: a systematic review of trials to identify features critical to success.” BMJ 330 (7494): 765. https://doi.org/10.1136/bmj.38398.500764.8F.
Kung, Tiffany H, Morgan Cheatham, Arielle Medenilla, Czarina Sillos, Lorie De Leon, Camille Elepaño, Maria Madriaga, et al. 2023. “Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models.” PLOS Digit Health 2 (2): e0000198. https://doi.org/10.1371/journal.pdig.0000198.
Ledley, R S, and L B Lusted. 1959. “Reasoning foundations of medical diagnosis; symbolic logic, probability, and value theory aid our understanding of how physicians reason.” Science 130 (3366): 9–21. https://doi.org/10.1126/science.130.3366.9.
Lin, Hause, and Tawab Safi. 2025. Ollamar: Ollama Language Models. https://hauselin.github.io/ollama-r/.
Lukac, Paul J, William Turner, Sitaram Vangala, Aaron T Chin, Joshua Khalili, Ya-Chen Tina Shih, Catherine Sarkisian, Eric M Cheng, and John N Mafi. 2025. “Ambient AI Scribes in Clinical Practice: A Randomized Trial.” NEJM AI 2 (12). https://doi.org/10.1056/aioa2501000.
McDuff, Daniel, Mike Schaekermann, Tao Tu, Anil Palepu, Amy Wang, Jake Garrison, Karan Singhal, et al. 2025. “Towards accurate differential diagnosis with large language models.” Nature 642 (8067): 451–57. https://doi.org/10.1038/s41586-025-08869-4.
McKinney, Scott Mayer, Marcin Sieniek, Varun Godbole, Jonathan Godwin, Natasha Antropova, Hutan Ashrafian, Trevor Back, et al. 2020. “International evaluation of an AI system for breast cancer screening.” Nature 577 (7788): 89–94. https://doi.org/10.1038/s41586-019-1799-6.
Mei, Xueyan, Hao-Chih Lee, Kai-Yue Diao, Mingqian Huang, Bin Lin, et al. 2020. “Artificial intelligence-enabled rapid diagnosis of patients with COVID-19.” Nat Med 26 (8): 1224–28. https://doi.org/10.1038/s41591-020-0931-3.
Miller, R A, H E Pople, and J D Myers. 1982. “Internist-1, an experimental computer-based diagnostic consultant for general internal medicine.” N Engl J Med 307 (8): 468–76. https://doi.org/10.1056/NEJM198208193070803.
Nivard, Michel, James Wade, and Samuel Calderon. 2025. Gptstudio: Use Large Language Models Directly in Your Development Environment. https://github.com/MichelNivard/gptstudio.
O’Sullivan, Jack W, Anil Palepu, Khaled Saab, Wei-Hung Weng, et al. 2026. “A large language model for complex cardiology care.” Nat Med 32 (2): 616–23. https://doi.org/10.1038/s41591-025-04190-9.
Omiye, Jesutofunmi A, Jenna C Lester, Simon Spichak, Veronica Rotemberg, and Roxana Daneshjou. 2023. “Large language models propagate race-based medicine.” NPJ Digit Med 6 (1): 195. https://doi.org/10.1038/s41746-023-00939-z.
OpenAI. 2022. “Introducing ChatGPT.” https://openai.com/index/chatgpt/.
———. 2023. “GPT-4 Technical Report.” https://arxiv.org/abs/2303.08774.
Ouyang, Long, Jeff Wu, Xu Jiang, Diogo Almeida, Carroll L. Wainwright, Pamela Mishkin, et al. 2022. “Training Language Models to Follow Instructions with Human Feedback.” In Advances in Neural Information Processing Systems 35 (NeurIPS 2022). https://arxiv.org/abs/2203.02155.
R Consortium. 2024. “News from R Submissions Working Group – Pilot 3 Successfully Reviewed by FDA.” https://r-consortium.org/posts/news-from-r-submissions-working-group-pilot-3/.
Radford, Alec, Karthik Narasimhan, Tim Salimans, and Ilya Sutskever. 2018. “Improving Language Understanding by Generative Pre-Training.” Technical report. OpenAI.
Radford, Alec, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, and Ilya Sutskever. 2019. “Language Models Are Unsupervised Multitask Learners.” Technical report. OpenAI.
Rajpurkar, Pranav, Emma Chen, Oishi Banerjee, and Eric J Topol. 2022. “AI in health and medicine.” Nat Med 28 (1): 31–38. https://doi.org/10.1038/s41591-021-01614-0.
Roberts, Michael, Derek Driggs, Matthew Thorpe, Julian Gilbey, Michael Yeung, et al. 2021. “Common pitfalls and recommendations for using machine learning to detect and prognosticate for COVID-19 using chest radiographs and CT scans.” Nat Mach Intell 3: 199–217. https://doi.org/10.1038/s42256-021-00307-0.
Ruiz, Edgar. 2025a. Chattr: Interact with Large Language Models in ’RStudio’. https://mlverse.github.io/chattr/.
———. 2025b. Mall: Run Multiple Large Language Model Predictions Against a Table, or Vectors. https://mlverse.github.io/mall/.
Sellergren, Andrew, Sahar Kazemzadeh, Tiam Jaroensri, Atilla Kiraly, Madeleine Traverse, et al. 2025. “MedGemma Technical Report.” https://arxiv.org/abs/2507.05201.
Shortliffe, Edward H. 1976. Computer-Based Medical Consultations: MYCIN. New York: Elsevier.
Singhal, Karan, Shekoofeh Azizi, Tao Tu, S Sara Mahdavi, Jason Wei, Hyung Won Chung, Nathan Scales, et al. 2023. “Large language models encode clinical knowledge.” Nature 620 (7972): 172–80. https://doi.org/10.1038/s41586-023-06291-2.
Singhal, Karan, Tao Tu, Juraj Gottweis, Rory Sayres, Ellery Wulczyn, Mohamed Amin, Le Hou, et al. 2025. “Toward expert-level medical question answering with large language models.” Nat Med 31 (3): 943–50. https://doi.org/10.1038/s41591-024-03423-7.
Thirunavukarasu, Arun James, Darren Shu Jeng Ting, Kabilan Elangovan, Laura Gutierrez, Ting Fang Tan, and Daniel Shu Wei Ting. 2023. “Large language models in medicine.” Nat Med 29 (8): 1930–40. https://doi.org/10.1038/s41591-023-02448-8.
Topol, Eric J. 2019. “High-performance medicine: the convergence of human and artificial intelligence.” Nat Med 25 (1): 44–56. https://doi.org/10.1038/s41591-018-0300-7.
Touvron, Hugo, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, et al. 2023. “LLaMA: Open and Efficient Foundation Language Models.” https://arxiv.org/abs/2302.13971.
Tu, Tao, Mike Schaekermann, Anil Palepu, Khaled Saab, Jan Freyberg, Ryutaro Tanno, Amy Wang, et al. 2025. “Towards conversational diagnostic artificial intelligence.” Nature 642 (8067): 442–50. https://doi.org/10.1038/s41586-025-08866-7.
Van Veen, Dave, Cara Van Uden, Louis Blankemeier, Jean-Benoit Delbrouck, Asad Aali, Christian Bluethgen, Anuj Pareek, et al. 2024. “Adapted large language models can outperform medical experts in clinical text summarization.” Nat Med 30 (4): 1134–42. https://doi.org/10.1038/s41591-024-02855-5.
Vaswani, Ashish, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. “Attention Is All You Need.” In Advances in Neural Information Processing Systems 30 (NeurIPS 2017). https://arxiv.org/abs/1706.03762.
Wang, Dandan, and Shiqing Zhang. 2024. “Large Language Models in Medical and Healthcare Fields: Applications, Advances, and Challenges.” Artificial Intelligence Review 57 (11): 299. https://doi.org/10.1007/s10462-024-10921-0.
Wickham, Hadley, Joe Cheng, Aaron Jacobs, Garrick Aden-Buie, and Barret Schloerke. 2026. Ellmer: Chat with Large Language Models. https://ellmer.tidyverse.org.
Wong, Andrew, Erkin Otles, John P Donnelly, Andrew Krumm, Jeffrey McCullough, Olivia DeTroyer-Cooley, Justin Pestrue, et al. 2021. “External Validation of a Widely Implemented Proprietary Sepsis Prediction Model in Hospitalized Patients.” JAMA Intern Med 181 (8): 1065–70. https://doi.org/10.1001/jamainternmed.2021.2626.
Wynants, Laure, Ben Van Calster, Gary S Collins, Richard D Riley, Georg Heinze, et al. 2020. “Prediction models for diagnosis and prognosis of covid-19: systematic review and critical appraisal.” BMJ 369: m1328. https://doi.org/10.1136/bmj.m1328.

Thank you!

Dr. Wan Nor Arifin

wnarifin@usm.my · wnarifin.github.io

Slides: wnarifin.github.io/2026-r-conf