This week’s AI in Healthcare and Digital Health update covers advances across clinical AI and diagnostics, including regulatory steps, Phase 3 trial design choices, large real-world data readouts, and cross-sector partnerships that shape care delivery and screening pathways.

In Today’s Newsletter

Dive deeper

In Today’s Newsletter

👁️‍🧠 NeuraLight biomarkers join TT-P34 Parkinson’s trial [1] [21 Jan 2026]

https://aijourn.com/neuralight-and-teitur-trophics-announce-collaboration-to-use-precision-biomarkers-in-tt-p34-parkinsons-disease-clinical-trial/

Context: Teitur Trophics developing TT-P34, a SorCS2-derived peptide with neuroprotective mechanism.

Key point: Teitur gets access to NeuraLight’s eye-movement brain-function biomarkers for the upcoming PD clinical trial.

Implication: Signals pipeline investment and modality expansion.

🫁 Brainomix e-Lung set as co-primary endpoint in BI Phase 3b ILD study [2] [20 Jan 2026]

https://www.prnewswire.com/news-releases/brainomix-e-lung-ai-imaging-technology-selected-as-co-primary-endpoint-in-boehringer-ingelheim-phase-3-clinical-trial-in-pulmonary-fibrosis-302664595.html

Context: DROP-FPF will assess nerandomilast (Jascayd) in high-risk individuals, with 2-year follow-up.

Key point: First Phase 3 ILD trial to use automated HRCT AI biomarkers as co-primary endpoint.

Implication: May influence prescriber choice and payer reviews pending full data.

🫁 Qure.ai wins Gates funding for open lung datasets and POCUS AI [3] [22 Jan 2026]

https://www.mobihealthnews.com/news/asia/qureai-scores-multimillion-dollar-gates-grant-ai-lung-ultrasound-and-more-briefs

Context: Funds to build multimodal datasets aligned to WHO lung pathways and develop POCUS algorithms.

Key point: Grant supports early TB and pneumonia detection tools in under-resourced settings.

Implication: Access programs may expand screening, initiation, and follow-up at scale.

🔬 fast-RSOM skin imaging flags microvascular dysfunction for CVD risk [4] [23 Jan 2026]

https://www.digitalhealthnews.com/skin-imaging-technology-now-detects-heart-disease-warning-signs

Context: German teams used fast-RSOM to visualize capillary-level function in smokers and CVD patients.

Key point: Noninvasive scans detected early endothelial dysfunction before structural changes.

Implication: Could inform practice and payer discussions; interpretation depends on study design and confounding control.

🧫 Cancilico raises €2.5M for AI bone marrow diagnostics MyeloAID [5] [20 Jan 2026]

https://pulse2.com/cancilico-e2-5-million-seed-funding-closed-to-scale-ai-powered-bone-marrow-diagnostics/

Context: Dresden startup with RUO software via PathoZoom; pursuing FDA and CE-IVDR.

Key point: Seed round to scale commercialization and digital biomarker development for hematologic malignancies.

Implication: Signals pipeline investment and modality expansion.

👶 Citizen Health backs DiMe DATAcc pediatric rare disease measures [6] [22 Jan 2026]

https://www.tipranks.com/news/private-companies/citizen-health-backs-dime-datacc-digital-measures-initiative-for-pediatric-rare-diseases

Context: Open-access digital measures project highlighted in a webinar.

Key point: Company supports standards to speed research and care for pediatric rare diseases.

Implication: Access programs may expand screening, initiation, and follow-up at scale.

👁️ Remidio DR AI shows high real-world accuracy in NHS-led study [7] [25 Jan 2026]

https://vohnetwork.com/amp/story/news/healthtech/remidios-ai-shows-strong-real-world-accuracy-in-diabetic-retinopathy-study

Context: Independent evaluation across 200k screening encounters, ~1.2M images.

Key point: Reported strong sensitivity for referable and sight-threatening DR while maintaining specificity.

Implication: May influence prescriber choice and payer reviews pending full data.

💊 Meta-analysis: DHIs did not improve medication adherence overall [8] [23 Jan 2026]

https://journals.sagepub.com/doi/10.1177/20552076251361593

Context: 13 studies, N=1320; subgrouped by design, duration, platform.

Key point: No significant adherence benefit vs usual care; signal only in 6-month studies.

Implication: Could inform practice and payer discussions; interpretation depends on study design and confounding control.

🤳 Mobile measles self-assessment tool shows high accuracy [9] [19 Jan 2026]

https://bmjdigitalhealth.bmj.com/content/2/1/e000082

Context: Deep learning on 44k+ images plus symptom questionnaire; external validations reported.

Key point: Tool achieved high accuracy and is intended for low-resource screening and triage.

Implication: Access programs may expand screening, initiation, and follow-up at scale.

🚴‍♀️ Co-design boosts adherence to VR physio-cognitive intervention [10] [24 Jan 2026]

https://www.nature.com/articles/s41746-026-02351-9

Context: Frail nursing home residents co-designed gamified modules for a stationary-bike VR.

Key point: Reported increases in duration and retention adherence plus psychological benefits.

Implication: Could inform practice and payer discussions; interpretation depends on study design and confounding control.

👁️‍🗨️ AI for glaucoma progression: review calls for standards [11] [22 Jan 2026]

https://www.nature.com/articles/s41746-026-02372-4

Context: 43 studies since 2014 assessed with QUADAS-2.

Key point: Moderate–good performance, but heterogeneity and generalisability gaps limit translation.

Implication: Could inform practice and payer discussions; interpretation depends on study design and confounding control.

🦴 SpAgents multi-agent system aids early axSpA diagnosis [12] [22 Jan 2026]

https://www.nature.com/articles/s41746-026-02372-4

Context: LLM+imaging agents validated on 596 patients across hospitals.

Key point: Higher sensitivity and accuracy than PCPs and juniors; comparable to seniors in testing.

Implication: Could inform practice and payer discussions; interpretation depends on study design and confounding control.

🧠 NEJM AI: silent real-time DCI prediction reduced false positives [13] [22 Jan 2026]

https://ai.nejm.org/doi/full/10.1056/AIoa2500749

Context: 63 SAH patients; 794 assessments; algorithm outputs reviewed without affecting care.

Key point: Fewer false positives and fewer intensified neuro checks without missing DCI events.

Implication: Could inform practice and payer discussions; interpretation depends on study design and confounding control.

🫀 TARGET-AI framework for targeted AI-ECG screening [14] [22 Jan 2026]

https://ai.nejm.org/doi/full/10.1056/AIoa2500588

Context: EHR foundation model plus contrastive ECG-echo model; validated across cohorts.

Key point: Detected 26 SHD subtypes; targeted deployment reduced false positives and improved F1 vs untargeted.

Implication: Could inform practice and payer discussions; interpretation depends on study design and confounding control.

🗣️ Wearable “intelligent throat” restores natural speech in stroke [15] [19 Jan 2026]

https://www.nature.com/articles/s41467-025-68228-9

Context: Textile sensors plus LLM agents for token-level decoding in five dysarthric patients.

Key point: Low word and sentence error rates with higher user satisfaction in pilot.

Implication: Could inform practice and payer discussions; interpretation depends on study design and confounding control.

🧩 OpenAI CFO on scaling with the value of intelligence [16] [18 Jan 2026]

https://openai.com/index/a-business-that-scales-with-the-value-of-intelligence/

Context: Company narrative on compute, product adoption, and monetization.

Key point: Reported ARR growth alongside compute scale and plans for agents and workflow automation.

Implication: Signals pipeline investment and modality expansion.

🧪 NeurIPS 2025 audit finds fabricated citations in accepted papers [17] [22 Jan 2026]

https://the-decoder.com/over-100-fake-citations-slip-through-peer-review-at-top-ai-conference/

Context: GPTZero analyzed 4,841 accepted papers.

Key point: At least 100 hallucinated citations across 51 papers despite peer review.

Implication: Could inform practice and payer discussions; interpretation depends on study design and confounding control.

🧬 Insilico’s AI-designed NLRP3 inhibitor IND cleared for PD [18] [23 Jan 2026]

https://www.news-medical.net/news/20260123/AI-designed-NLRP3-inhibitor-receives-FDA-clearance-for-Parkinson-disease-trials.aspx

Context: ISM8969 to enter Phase 1 in healthy volunteers; co-development with Hygtia.

Key point: IND clearance for orally available, brain-penetrant NLRP3 inhibitor designed with AI.

Implication: May influence prescriber choice and payer reviews pending full data.

🏛️ ARPA-H ADVOCATE targets agentic AI for heart failure care [19] [21 Jan 2026]

https://www.fiercehealthcare.com/ai-and-machine-learning/trump-administration-creating-clinical-ai-agents-3-year-fda-approval

Context: Seeks FDA-authorized generative AI assistant and a supervisory agent; selection phases planned.

Key point: Aims for 24/7 patient-facing cardiovascular agents integrated with EHR and wearables.

Implication: Signals pipeline investment and modality expansion.

🫁 BMS–Microsoft partner on AI lung cancer early detection [20] [20 Jan 2026]

https://news.bms.com/news/corporate-financial/2026/Bristol-Myers-Squibb-Announces-Collaboration-with-Microsoft-to-Advance-AI-Driven-Early-Detection-of-Lung-Cancer/default.aspx

Context: Deploy FDA-cleared radiology AI via Microsoft Precision Imaging Network across U.S. sites.

Key point: Collaboration aims to surface nodules earlier and improve follow-up pathways, with equity focus.

Implication: Access programs may expand screening, initiation, and follow-up at scale.

Why it matters

  • Endpoint selection by BI validates quantitative imaging AI as trial-grade measure in ILD.
  • Real-world DR AI performance at scale supports national screening workflows.
  • Mixed evidence on DHIs for adherence argues for longer-duration, rigorously designed interventions.
  • Foundation models plus targeting can curb false positives and clinician burden in cardiology AI.
  • Regulatory and governance moves, from ARPA-H to INDs, signal routes for generative and discovery AI into clinics.

🚀 Accelerate your success. Contact us now

📂 Explore our case studies. See examples of our work.

💡 Read our insights. Learn from our latest reports and analysis

🎬 Watch on YouTube. Subscribe and never miss a video.

🧰 See our full range of servicesDiscover how we can help you.

📚 Find your one-stop page for the full AI and Digital Health archive.

FAQ

What is TT-P34 and how is NeuraLight involved? [1]

TT-P34 is a SorCS2-derived peptide in development for Parkinson’s. Teitur will use NeuraLight’s eye-movement biomarkers in its upcoming trial to quantify brain function effects, endpoint not specified.

How will Brainomix e-Lung be used in DROP-FPF? [2]

As a co-primary endpoint, the FDA-cleared HRCT AI will quantify fibrosis extent and severity over time in a Phase 3b study of nerandomilast in ILD risk populations.

What did the NHS-led DR AI evaluation show for Remidio? [7]

Across 200k screenings, Remidio’s CE-marked AI had high sensitivity for referable and sight-threatening DR while maintaining specificity and predictive value, supporting safe referral and operational efficiency.

Did digital health interventions improve medication adherence? [8]

Overall, no. A meta-analysis of 13 studies found no significant benefit vs usual care, with a positive signal only in 6-month interventions; endpoints varied and heterogeneity was moderate.

How does TARGET-AI reduce AI-ECG false positives? [14]

By targeting screening using EHR-derived phenotypes and a contrastive ECG–echo model, it reduced false positives and improved F1 across multiple structural heart disease labels versus untargeted strategies.

What is ARPA-H’s ADVOCATE program aiming to approve? [19]

An FDA-authorized agentic AI assistant for cardiovascular care with defined intended use in heart failure and a disease-agnostic supervisory agent for ongoing oversight; deployment and down-select phases are planned.

Entities / Keywords

NeuraLight; Teitur Trophics; TT-P34; Parkinson’s disease; Brainomix e-Lung; Boehringer Ingelheim; HRCT; ILD; Qure.ai; Gates Foundation; fast-RSOM; Cancilico MyeloAID; DiMe DATAcc; Citizen Health; Remidio DR AI; medication adherence DHIs; Mobile measles AI tool; VR physio-cognitive intervention; glaucoma AI progression; SpAgents axSpA; NEJM AI DCI model; TARGET-AI ECG; Intelligent Throat wearable; OpenAI business model; NeurIPS citations; Insilico ISM8969 NLRP3; ARPA-H ADVOCATE; Bristol Myers Squibb–Microsoft Precision Imaging Network.

References

  1. https://aijourn.com/neuralight-and-teitur-trophics-announce-collaboration-to-use-precision-biomarkers-in-tt-p34-parkinsons-disease-clinical-trial/
  2. https://www.prnewswire.com/news-releases/brainomix-e-lung-ai-imaging-technology-selected-as-co-primary-endpoint-in-boehringer-ingelheim-phase-3-clinical-trial-in-pulmonary-fibrosis-302664595.html
  3. https://www.mobihealthnews.com/news/asia/qureai-scores-multimillion-dollar-gates-grant-ai-lung-ultrasound-and-more-briefs
  4. https://www.digitalhealthnews.com/skin-imaging-technology-now-detects-heart-disease-warning-signs
  5. https://pulse2.com/cancilico-e2-5-million-seed-funding-closed-to-scale-ai-powered-bone-marrow-diagnostics/
  6. https://www.tipranks.com/news/private-companies/citizen-health-backs-dime-datacc-digital-measures-initiative-for-pediatric-rare-diseases
  7. https://vohnetwork.com/amp/story/news/healthtech/remidios-ai-shows-strong-real-world-accuracy-in-diabetic-retinopathy-study
  8. https://journals.sagepub.com/doi/10.1177/20552076251361593
  9. https://bmjdigitalhealth.bmj.com/content/2/1/e000082
  10. https://www.nature.com/articles/s41746-026-02351-9
  11. https://www.nature.com/articles/s41746-026-02372-4
  12. https://www.nature.com/articles/s41746-026-02372-4
  13. https://ai.nejm.org/doi/full/10.1056/AIoa2500749
  14. https://ai.nejm.org/doi/full/10.1056/AIoa2500588
  15. https://www.nature.com/articles/s41467-025-68228-9
  16. https://openai.com/index/a-business-that-scales-with-the-value-of-intelligence/
  17. https://the-decoder.com/over-100-fake-citations-slip-through-peer-review-at-top-ai-conference/
  18. https://www.news-medical.net/news/20260123/AI-designed-NLRP3-inhibitor-receives-FDA-clearance-for-Parkinson-disease-trials.aspx
  19. https://www.fiercehealthcare.com/ai-and-machine-learning/trump-administration-creating-clinical-ai-agents-3-year-fda-approval
  20. https://news.bms.com/news/corporate-financial/2026/Bristol-Myers-Squibb-Announces-Collaboration-with-Microsoft-to-Advance-AI-Driven-Early-Detection-of-Lung-Cancer/default.aspx

Privacy Preference Center