Explainability and causability in digital pathology
Open accessAbstract
The current move towards digital pathology enables pathologists to use artificial intelligence (AI)-based computer programmes for the advanced analysis of whole slide images. However, currently, the best-performing AI algorithms for image analysis are deemed black boxes since it remains - even to their developers - often unclear why the algorithm delivered a particular result. Especially in medicine, a better understanding of algorithmic decisions is essential to avoid mistakes and adverse effects on patients. This review article aims to provide medical experts with insights on the issue of explainability in digital pathology. A short introduction to the relevant underlying core concepts of machine learning shall nurture the reader's understanding of why explainability is a specific issue in this field. Addressing this issue of explainability, the rapidly evolving research field of explainable AI (XAI) has developed many techniques and methods to make black-box machine-learning systems more transparent. These XAI methods are a first step towards making black-box AI systems understandable by humans. However, we argue that an explanation interface must complement these explainable models to make their results useful to human stakeholders and achieve a high level of causability, i.e. a high level of causal understanding by the user. This is especially relevant in the medical field since explainability and causability play a crucial role also for compliance with regulatory requirements. We conclude by promoting the need for novel user interfaces for AI applications in pathology, which enable contextual understanding and allow the medical expert to ask interactive 'what-if'-questions. In pathology, such user interfaces will not only be important to achieve a high level of causability. They will also be crucial for keeping the human-in-the-loop and bringing medical experts' experience and conceptual knowledge to AI processes.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
To be clinically useful, explanations must align with dermatopathological hallmarks such as epidermal changes and mitotic activity.
“To be clinically useful, explanations must align with dermatopathological hallmarks such as epidermal changes and mitotic activity [66].”
From Artificial Intelligence in Dermatopathology: An Update and Review of the Current Literature (2026) · cites this paper as [66]
The section on demonstrating scientific validity (page unavailable) says an AI algorithm should show that its Ki-67 result is based on stained tumour nuclei, making its prediction verifiable by a medical expert. This supports the general need for clinically meaningful explanations, but the paper does not specify dermatopathological hallmarks such as epidermal changes or mitotic activity. (LLM, conf=0.84)
Experimental — accuracy not yet measuredThe "Black Box" problem refers to the limited interpretability of deep learning models; pathologists sometimes cannot fully understand why an algorithm reaches a certain decision, which hinders clinical trust and accuracy.
“The “Black Box” problem refers to the limited interpretability of deep learning models; pathologists sometimes cannot fully understand why an algorithm reaches a certain decision, which hinders clinical trust and accuracy [66].”
From Artificial Intelligence in Dermatopathology: An Update and Review of the Current Literature (2026) · cites this paper as [66]
The Conclusion and future outlook describes deep-learning models as ‘black boxes’ whose decisions can be unclear even to AI experts. ‘From explainability to causability’ says explanations help pathologists decide whether to rely on an AI recommendation, and ‘Spurious correlations’ links unrecognized prediction failures to risks in clinical use. The paper does not say that limited interpretability itself reduces a model’s accuracy; it says opacity makes accuracy problems harder to identify. (LLM, conf=0.83)
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
The citation lists haven’t been fetched yet
References 48 · Cited by 99
Research works printed but not linked: 2
- [22]Plass M, Kargl M, Evans T, et al. Human‐AI interfaces are a central component of trustworthy AI. In: Explainable AI: Foundations, Methodologies and Applications. Intelligent Systems Reference Library, Volume 232, Mehta M, Palade V, Chatterjee I (Eds). Cham: Springer International Publishing, 2022; 225– 256.
- [34]Li O, Liu H, Chen C, et al. Deep learning for case‐based reasoning through prototypes: a neural network that explains its predictions. In: Proceedings of the AAAI Conference on Artificial Intelligence, Volume 32, McIlraith SA, Weinberger KQ (Eds). AAAI Press: Palo Alto, CA, 2018; 3530– 3537.
- Authors
- Plass, M., Kargl, M., Kiehl, T., Regitnig, P., Geißler, C., Evans, T., Zerbe, N., Carvalho, R., Holzinger, A., & Müller, H.
- Affiliations
- University of Graz (AT)
- Humboldt-Universität zu Berlin (DE)
- Technische Universität Berlin (DE)
- Corresponding author
- Müller, Heimo
- ORCID
- Venue
- The Journal of Pathology Clinical Research
- Publisher
- Wiley
- Published
- 2023-04-12
- Volume
- 9
- Issue
- 4
- Pages
- 251-260
- ISSN
- 2056-4538 (ISSN-L)
- PMC
- PMC10240147
- OpenAlex
- W4365137055
- Cited by
- 99
- References
- 48
- Citations by year
- 2026: 23 · 2025: 40 · 2024: 31 · 2023: 5
- FWCI
- 2.07
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- The Journal of Pathology Clinical Research (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- Austrian Science Fund (P‐32554)
- Horizon 2020 Framework Programme (824087, 826078, 857122, 874662)
- Österreichische Forschungsförderungsgesellschaft (879881)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Physical Sciences › Computer Science › Artificial Intelligence › AI in cancer detection
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Keywords
- explainable AI, digital pathology, whole slide images, black-box models, causality, explainability, user interface, human-in-the-loop, regulatory compliance, what-if questions, medical image analysis
- MeSH terms
- Pathologists, Algorithms, Artificial Intelligence, Humans, Image Processing, Computer-Assisted
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
