Advancing diagnostics with artificial intelligence
May 14, 2026
AI supports diagnostics through connecting data, enhancing clinical decision-making and improving laboratory workflows.
How will artificial intelligence transform diagnostics?
Healthcare systems and laboratories globally are under increasing pressure. They face financial strain1 and staff shortages2, while the rising prevalence of non-communicable diseases (NCDs)3 and continuous medical advancements constantly elevate expectations on the in vitro diagnostics (IVD) and healthcare industry. Rapid technological progress, particularly in data generation, is emerging to meet these challenges. Historically, however, disconnected systems and data silos have hampered efforts to leverage this vast data for optimizing healthcare systems and alleviating pressure.
Roche is committed to tackling these core health system challenges by building interoperable digital infrastructure in collaboration with laboratories and healthcare institutions. This digital infrastructure connects data from Roche and non-Roche instruments across laboratory disciplines with key information systems, such as the electronic health record (EHR) or laboratory information systems (LIS).
Artificial intelligence (AI) is a pivotal technology to meaningfully utilize this health system data at scale, enabling transformational advances. At Roche, we focus our AI efforts on three areas: prediction, performance and personalization. By integrating advanced technology with rigorous clinical standards, we are moving beyond traditional diagnostics toward a future of smarter healthcare, from early detection to customized patient care.
Predict: Early risk identification with LGI-Flag at Geisinger Health4
AI is key to transitioning healthcare from a focus on reactive treatment to proactive prediction. By utilizing advanced solutions like diagnostic and risk-stratifying algorithms, AI may help identify at-risk patients earlier. Such proactive detection can lead to better outcomes and higher survival rates for patients.
Case study: early disease detection with LGI-Flag at Geisinger Health4
LGI-Flag is a machine learning-based algorithm to aid in the detection of patients who are at increased risk of harboring colorectal cancer (CRC).
Challenge
How to identify patients overdue for CRC screening who are at risk.
US Preventive Service Task Force (USPTSF) screening guidelines recommend a colonoscopy every 10 years for adults aged 45–75 considered low to average risk.
Solution
Implementing LGI-Flag into Geisinger’s EHR and clinical workflow.4
*LGI-Flag is not a medical device in the U.S.
Detect more CRC cases
The flagged population showed 8× more CRC cases across all stages than routine CRC colonoscopy screening.⁴
Support screening completion
Among patients advised for colonoscopy, 70% completed screening, with significant findings reported.⁴
Prioritize limited resources
LGI Flag can help enrich screening populations where CRC screening resources are limited.⁴
Perform: Driving Operational Efficiency through AI
AI-powered tools are essential for automating administrative burdens and supporting health care professionals (HCPs) and laboratory staff. Opportunities range from optimizing workflows, simplifying data analysis, improving troubleshooting and preventative maintenance to supporting pathologists in their diagnosis with AI image analysis algorithms. Our goal is to streamline workflows so clinicians can focus on what matters most: the patient.
See how AI drives efficiency in tumor board preparation.
Accelerating troubleshooting with an AI service agent
Service AI agent is an AI-powered issue resolution chatbot seamlessly integrated into Roche’s Online Support. The agent is available 24/7 and can process images and text.
Challenge
Because clinical laboratories are vital to healthcare systems, problems with diagnostic systems can disrupt laboratory workflows. Fast troubleshooting is needed to help minimize downtime.
Solution
Roche has introduced an intelligent AI agent to enhance customer service. This smart agent is trained on a vast amount of data, including user manuals, method sheets and historical customer support records. The goal is to allow customers to resolve their issues as quickly as possible.
Learn how AI can drive operational efficiency in tumor board preparation.
Another example of AI in clinical workflows is tumor board preparation. A pilot study at Northside Hospital examined AI-powered clinical summarization and reported efficiency improvements, while identifying areas for further evaluation.
Personalize: Leveraging AI to Deliver Tailored Care
AI can help analyze large volumes of health data to support personalized predictions for patients and clinicians. These insights may help inform more tailored care, from advanced companion diagnostics (CDx) and digital pathology algorithms to personalized glucose predictions in continuous glucose monitoring (CGM).
Would you like to learn more?
Get in touch with our sales team.
Providing patients with personalized glucose insights
Accu-Chek® SmartGuide CGM Solution provides real-time glucose readings to help patients improve their glycemic control and time in range, while reducing glucose excursions with predictive features.*
Personalized predictive features:
30-minute low glucose predict: alerts patients when hypoglycemia is likely to happen within 30 minutes, so they can act immediately to avoid it.
2-hour glucose predict: shows patients their glucose level over the next 2 hours.
7-hour night low predict: shows patients their nocturnal hypoglycemia risk before bedtime, so they can take preventive action.
Patients can explore their glucose data further in the mySugr Glucose Insights app, supporting daily diabetes management beyond CGM readings.**
Nominated for the 2026 AI Impact Award in leveraging AI to make impact
Building on real-time glucose readings and predictive features, Roche’s AI-driven approach has been nominated for the 2026 AI Impact Award in the Product and Customer Experience category. In an interview, Dr Imane Moest, Head of Chronic Remote Care at Roche, explains how the approach was developed and how it is being applied in diabetes management.
- American College of Healthcare Executives. Navigating financial pressures in 2025 [Internet; cited 2025 Aug 25]. Available from: https://healthcareexecutive.org/sponsored-content/navigating-financial-pressures-in-2025
- World Economic Forum. How AI can transform patient care and treatment [Internet; updated 2024 Jan 17; cited 2024 Mar 29]. Available from: https://www.weforum.org/stories/2024/01/how-ai-can-transform-patient-care-and-treatment/
- World Health Organization. World health statistics 2023: monitoring health for the SDGs [PDF; cited 2024 Apr 7]. Available from: https://iris.who.int/bitstream/handle/10665/367912/9789240074323-eng.pdf
- Not every digital product is available in all markets. The use of any third-party app is subject to a separate license agreement with the respective third-party app developer. Roche gives no warranties (express or implied) with regard to any third-party app. Third-party apps might not be available in your country. This website and its content may be accessible worldwide, Roche assumes no liability with regard to the access to the information, which may not be compatible with legislations or regulations in force in your country.
- *The Accu-Chek SmartGuide Predict app predicts the general estimated glucose development within the next 2 hours (Glucose Predict), the estimated high risk of low glucose within the next 30 minutes (Low Glucose Predict) and the estimated risk of a night-time hypo for the upcoming 7 hours (Night Low Predict, active between 21:00 and 02:00).
- **mySugr Glucose Insights is not available in all markets.