From Data to Diagnosis: How AI in Pathology Supports Better Outcomes

In this article, we’ll explore how AI-powered tools are beginning to transform immunoassay cancer diagnostics for the better.

Key takeaways:

• Early and accurate cancer diagnosis is essential, but interpreting oncology marker immunoassays is often difficult due to imperfect specificity and sensitivity or borderline results.

• Human judgment of immunoassay results,  especially those with subtle or ambiguous findings, can vary between doctors. This leads to potential errors and delays.

• AI and ML are increasingly embedded in diagnostic analyzers and lab software, allowing analysis of complex patterns rather than traditional thresholds.

• AI can assist or outperform human pathologists, reducing observer variability and improving agreement, accuracy and efficiency across sites.

• Using AI in pathology speeds up result turnaround, helps detect cancer or recurrence earlier, and enables more standardized interpretations regardless of location.

• Data quality still matters. Even the most advanced AI relies on high-quality immunoassays and reagents. This is where Medix Biochemica can assist.

Did you know that one in four deaths in the United States is due to cancer?¹ This means that early and accurate detection is more important than ever before, but interpreting oncology marker tests is often a complex process.

Immunoassays – the tests that use antibodies to detect cancer-related proteins in blood, tissue or other samples – are the foundation of cancer detection and monitoring.^1,2

Within modern pathology, including the growing use of AI in pathology, manually judging immunoassay results isn’t always straightforward. Immunoassays have multiple known limitations as they interact with human samples, so results aren’t always a clear “yes” or “no”.³ This is why artificial intelligence (AI) is now being used to assist. AI cancer detection technologies are being embedded in diagnostic analyzers and lab software, promising to make immunoassay-based cancer testing faster, more reliable, more personalized and more impactful for patients.^4,5

Why is oncology marker interpretation so complex?

Oncology markers are substances (usually proteins) that can be measured by immunoassays to give clues about cancer. These tests are valuable, but subtle changes or borderline levels can be difficult to read. 

For example, a blood test might come back only slightly elevated – high enough to raise concern, but not high enough to be a definitive positive.⁶ Doctors then often need to order repeat tests or use other clinical information to avoid false positives and understand ambiguous results.⁶

No oncology marker test is perfect on its own. Most markers lack ideal sensitivity or specificity, so results must be interpreted alongside scans, biopsies and the patient’s medical history. This process is highly dependent on the doctor’s experience and judgment.⁷

Interpreting manual cancer test results: What are the challenges?

• Observer variability/variation: Different doctors (or even the same doctor on different days) can read results differently,⁸ especially when these results are borderline.

• Immunoassays often yield results that aren’t clearly normal or abnormal, but somewhere in between.⁶ These subtle signals are notoriously hard to read reliably: minor elevations in a marker might stem from benign conditions (e.g. the ovarian cancer marker CA-125 can rise due to pregnancy or inflammation).⁶ On the other hand, early-stage cancers might only cause a slight uptick that immunoassays are not specific or sensitive enough to identify.⁹

• Manual interpretation demands expertise, and even then it’s prone to errors or delays, causing stress and uncertainty for both doctors and patients awaiting answers.^6-9

This is where the use of AI in pathology can help.

AI cancer detection: AI integration in immunoassay analyzers

To address the challenges of interpreting cancer immunoassays, diagnostic companies are increasingly building AI and machine learning (ML) into immunoassay analyzers and lab software. Unlike a human, who might apply a single threshold, AI systems can analyze complex patterns in the data.⁹

AI and ML can analyze patient data, test histories and current symptoms to guide doctors in selecting the most appropriate diagnostic tests, reducing unnecessary investigations and saving on costs.⁹

Rather than simply flagging an abnormal result, ML algorithms can correlate findings with known disease trajectories, enabling more accurate and timely diagnoses.⁹

Demonstrating how AI in pathology is a promising tool for enhancing accurate test interpretation and reducing observer variability,¹⁰ a recent study showed that AI improved accuracy, agreement and efficiency among pathologists for Ki67 assessments in breast cancer.¹¹

Read the full study

Certain AI cancer detection tools are already outperforming human pathologists. One AI-based tool for analyzing PD-L1 immunohistochemistry in cancer biopsies achieved about 88% accuracy compared to 75% for manual scoring, and it was more sensitive in catching low-level positive cases. The tech performed consistently across many samples and even across different hospital sites.¹²

AI assistance is reducing error rates and tightening the range of interpretations between different users,¹¹ meaning more reproducible results and clearer decisions. A doctor in one city and another halfway around the world should be able to interpret a tricky immunoassay result the same way when they’re supported by the same well-trained AI cancer detection system.

What are the benefits of AI cancer detection for patients?

• Faster turnaround of results: by automating parts of analysis, AI can significantly speed up the diagnostic process. Tasks that once waited for a specialist’s manual review can be done in seconds by a machine.⁹ Faster results mean patients spend less time anxiously waiting and can start treatment earlier when needed.

• Earlier detection and better monitoring: AI’s pattern-recognition abilities allow it to catch important signals earlier than traditional testing methods.¹³ By sifting through subtle changes across multiple biomarkers and comparing against vast datasets, AI might detect the faint fingerprint of cancer recurrence or progression sooner. Catching cancer or relapse early on, when it’s more treatable, directly improves patient outcomes.¹⁴

• Consistent outcomes across testing sites: because an AI algorithm applies uniform criteria, it reduces variations in result interpretation across multiple testing sites.¹² As AI in pathology becomes more widespread, we can expect more standardization, meaning a patient in a small clinic can receive the same quality of test interpretation as they would at a major cancer center. 

In short, AI is helping ensure that wherever patients go, their immunoassay results will be quick, reliable and comparable, a critical step for quality care.

Frequently Asked Questions

Can AI detect tumors? 
Yes, AI and ML can be used in oncology to detect tumors with high accuracy across various cancer types.¹⁵

How does AI read biomarker thresholds?
AI interprets biomarker thresholds by analyzing complex biological data to identify patterns that define normal, abnormal or predictive states.¹⁶ Unlike traditional methods that rely on fixed, population-based reference ranges, AI models can use ML and deep learning to detect subtle relationships across thousands of biomarkers simultaneously.¹⁶

What are the benefits of AI for cancer detection?
Using AI for cancer detection can lead to faster turnaround of test results, earlier diagnosis, more effective monitoring and improved consistency in results across multiple sites.

How can Medix Biochemica help?
While AI in oncology is poised to revolutionize the diagnostic field, it’s important to remember that any diagnostic test is only as good as the assay it relies on. AI can analyze data and spot patterns, but the data must be accurate and specific in the first place. That means the quality of core reagents is non-negotiable.¹⁷

Medix Biochemica provides the high-quality raw materials needed to build sensitive and specific immunoassays, including those used for cancer detection. Get in touch with our experienced team to discuss how our expertise can support your success.¹⁷

Have you subscribed to our blog yet? Subscribe to our articles for the latest news and trends in the in vitro diagnostics field

References

1. Meany DL, Sokoll LJ, Chan DW. Early detection of cancer: Immunoassays for plasma tumor markers. Expert Opin Med Diagn. 2009;3(6):597-605. doi:10.1517/17530050903266830.
2. Wu J, Fu Z, Yan F, et al. Biomedical and clinical applications of immunoassays and immunosensors for tumor markers. TrAC Trends in Analytical Chemistry. 2007;26(7):679-688. doi:10.1016/j.trac.2007.05.007.
3. Hoofnagle AN, Wener MH. The fundamental flaws of immunoassays and potential solutions using tandem mass spectrometry. J Immunol Methods. 2009;347(1-2):3-11. doi:10.1016/j.jim.2009.06.003.
4. Tiwari A, Mishra S, Kuo TR. Current AI technologies in cancer diagnostics and treatment. Mol Cancer. 2025;24(1):159. doi:10.1186/s12943-025-02369-9.
5. AI-generated sensors open new paths for early cancer detection. MIT News. Accessed January 24, 2026. https://news.mit.edu/2026/ai-generated-sensors-open-new-paths-early-cancer-detection-0106.
6. The key ovarian cancer blood test for detection. Liv Hospital. Accessed January 24, 2026. https://int.livhospital.com/the-key-ovarian-cancer-blood-test-for-detection/.
7. Mahadevarao Premnath S, Zubair M. Laboratory evaluation of tumor biomarkers. In: StatPearls. Accessed January 24, 2026. http://www.ncbi.nlm.nih.gov/books/NBK597378/.
8. Observer variation. UMass Profiles. Accessed January 26, 2026. https://profiles.umassmed.edu/display/113672.
9. Revolutionizing the lab with AI in laboratory workflows. Diagnostics. Accessed January 26, 2026. https://diagnostics.roche.com/global/en/lab-leaders/article/machine-learning-ai-in-laboratory.html.
10. Tizhoosh HR, Diamandis P, Campbell CJV, et al. Searching images for consensus: Can AI remove observer variability in pathology? Am J Pathol. 2021;191(10):1702-1708. doi:10.1016/j.ajpath.2021.01.015.
11. Dy A, Nguyen NNJ, Meyer J, et al. AI improves accuracy, agreement and efficiency of pathologists for Ki67 assessments in breast cancer. Sci Rep. 2024;14:1283. doi:10.1038/s41598-024-51723-2.
12. AI tool outperforms manual PD-L1 scoring in a study. ESMO Daily Reporter. Accessed January 26, 2026. https://dailyreporter.esmo.org/news/ai-tool-outperforms-manual-pd-l1-scoring-in-a-study.
13. Hunter B, Hindocha S, Lee RW. The role of artificial intelligence in early cancer diagnosis. Cancers (Basel). 2022;14(6):1524. doi:10.3390/cancers14061524.
14. Zhang B, Shi H, Wang H. Machine learning and AI in cancer prognosis, prediction, and treatment selection: A critical approach. J Multidiscip Healthc. 2023;16:1779-1791. doi:10.2147/JMDH.S410301.
15. Pesheva E. New AI tool can diagnose cancer, guide treatment, predict patient survival. Harvard Gazette. Accessed January 26, 2026. https://news.harvard.edu/gazette/story/2024/09/new-ai-tool-can-diagnose-cancer-guide-treatment-predict-patient-survival/.
16. Composite intelligence: How AI-driven biomarkers are setting the bar in immuno-oncology. The Cancer Letter. Accessed January 26, 2026. https://cancerletter.com/sponsored-article/20260109_6/.
17. The benefits of AI in endocrinology. Medix Biochemica. Accessed January 26, 2026. https://articles.medixbiochemica.com/the-benefits-of-ai-in-endocrinology-medix-biochemica.