The study, published in the International Journal of Computer Assisted Radiology and Surgery, is a feasibility experiment rather than a clinical trial, but its implications reach far beyond its modest size. The researchers asked a deceptively simple question: if a hospital information system recorded every interaction among patients, staff and devices as timestamped, spatially anchored, context-rich events, would the resulting data be directly usable for the modern analytical methods that healthcare research increasingly depends on, such as process mining and machine learning? Their answer, based on a simulated pre-operative workflow involving eighteen volunteers, is a qualified yes.

The conceptual foundation comes from earlier work by the same group, which proposed OMNI-SYS as a first step toward an object-oriented hospital information system, abbreviated oHIS. Unlike traditional systems that store data in disconnected relational tables, an oHIS represents patients, nurses, doctors, devices and rooms as objects with properties and relationships that evolve across three dimensions. The term pragmatic is deliberate: the framework captures only the data relevant to the process at hand, which keeps it flexible and scalable, whether information is entered manually, scanned via QR codes, or gathered through advanced techniques such as computer vision. This design choice allows the system to function as a living digital twin, dynamically mapping objects and updating their states in real time.

To test the concept under realistic conditions, the team conducted the study at TUM University Hospital in December 2025. For ethical reasons, they recruited eighteen healthy volunteers instead of real patients, a practice consistent with established healthcare simulation research in which standardized patients are used to analyze patient flow and system usability. The cohort size was chosen deliberately to approximate the maximum number of gastrointestinal surgery patients undergoing pre-operative testing at the local surgical department on a single clinical day. Each volunteer was assigned one of seven common gastrointestinal procedures and followed a predefined pre-operative pathway through real hospital departments, including administrative registration, premedication consultation, electrocardiography, lung function testing, endoscopy, computed tomography and magnetic resonance imaging.

The pathways were designed by an interdisciplinary group of surgeons, data scientists and engineers, following the Surgineering paradigm of close clinical-technical collaboration. Every pathway began and ended at the administrative registration office, which serves in the institution’s real workflow as both the entry and exit point where patients hand over collected documents. While the set of required stations was fixed for each procedure, the order of visits was intentionally varied so that two patients undergoing the same operation could move through the hospital in different sequences, loosely reflecting the variability of genuine clinical settings. Each participant received a personalized schedule, and upon arrival at each station their state was updated in the system.

Data collection relied on a web-based patient tracking interface integrated into OMNI-SYS, which allowed researchers to define pathways, assign patient-specific sequences and monitor progress in real time. Each station carried a QR code that participants scanned with smartphones upon arrival, generating timestamped events against a central server clock. As a redundancy measure, a researcher at each station also recorded arrival times manually, and all timestamps were standardized to one-minute resolution for analysis. The protocol specified that manually collected timestamps would be used whenever the two methods disagreed by more than sixty seconds, though no such discrepancy arose. Over the course of a thirty-minute session, the eighteen participants generated eighty-eight timestamped events, with most completing their pathways within about seven minutes and a median journey time of roughly five minutes.

The analytical core of the study examined whether this object-centric event data could feed directly into contemporary process mining tools. Using the PM4Py library, the researchers constructed a directly-follows graph in which activities corresponded to patient locations and arcs showed transitions with average travel times. The graph captured multiple pathway variants, showing for instance that patients could move from administrative registration to five different locations. Because participants did not undergo actual clinical procedures, activity durations registered as effectively zero, and the model represented prescribed routes rather than spontaneous behavior. Applied to real patients, the authors note, the same approach could capture work-as-done rather than work-as-imagined, a distinction that process mining researchers consider fundamental.

More revealing was the object-centric extension. When the event log was converted to the Object-Centric Event Log format, with patients, devices, nurses and doctors declared as object types, the resulting object-centric directly-follows graph could model the lifecycles of all agent types simultaneously. Patient edges traced movement between stations, while stationary resource objects produced edges and self-loops indicating continuous reuse, such as a device serving successive patients. The graph clearly separated two patterns: patients flowing through the hospital and resources being shared across concurrent cases. Because the pathways were designed so that all procedures passed through administrative registration twice while only one included lung function testing, the framework correctly displayed higher nurse engagement at the former, demonstrating that it captures resource utilization and sharing across overlapping patient journeys.

Network analysis added a further layer of insight. A transition graph built with Python’s NetworkX library revealed a hub-and-spoke structure centered on the triangular core of administrative registration, ECG and premedication stations, where the strongest patient flows occurred. The researchers also quantified patient co-occurrence, defined as multiple patients sharing the same location, time and context within one-minute windows. Administrative registration showed high-frequency co-occurrences concentrated early in the session, while specialized stations like MRI and lung function showed none. The authors are careful to stress that in this simulated setting such patterns illustrate what the data structure could reveal rather than proving congestion, since procedure and waiting times were not recorded. Even so, in a real deployment this kind of measurement could flag locations at risk of bottlenecks before they become visible to staff.

The study’s limitations are candidly acknowledged. The dataset is too small for meaningful machine learning, the pathways were predetermined rather than observed, and any process model fitted to them would risk overfitting, so the process mining results stand as proof of concept rather than validated models. QR-code scanning temporarily failed during parts of the session due to network issues, and manual backups kept collection going, underscoring the need for robust multi-modal tracking in future work. Yet the structural achievement is clear: the object-centric data can be replayed as a scenario within OMNI-SYS, supports conformance checking and resource engagement analysis, and offers the consistent feature definitions that machine learning depends on. The authors argue that following a few hundred real patients would likely suffice for predictive modeling, and that pairing spatiotemporal forecasting with optimization solvers or reinforcement learning could eventually improve staff scheduling, appointment routing and congestion management. As a blueprint for object-centric clinical research in surgery, the framework points toward hospital information systems that no longer fragment reality but mirror it, bridging the persistent gap between clinical routine and research.

Subject of Research: A pragmatic object-centric hospital digital twin that records patient, staff and device interactions across location, time and context to enable process mining and workflow optimization.

Article Title: Expanding research on clinical agent data to a virtual space of location, time, and context: exploration of data structures and limitations using a pragmatic digital twin

Article References: Rashid, S., Sliepkova, K., Bernhard, L., Stabenow, S., Spicker, E., Rinderle-Ma, S., Fottner, J., Wilhelm, D., & Berlet, M. (2026). Expanding research on clinical agent data to a virtual space of location, time, and context: exploration of data structures and limitations using a pragmatic digital twin. International Journal of Computer Assisted Radiology and Surgery. https://doi.org/10.1007/s11548-026-03788-9

Image Credits: AI Generated

DOI: 10.1007/s11548-026-03788-9

Keywords: digital twin, hospital information system, process mining, object-centric data, clinical workflow, pre-operative pathway, patient tracking, health informatics, machine learning, data integration, workflow optimization, OMNI-SYS

The problem, the researchers argue, is that most existing similarity and difference analysis techniques rely on relatively single analysis criteria. Traditional methods typically consider only the dependency relations between activities in a process model. That works reasonably well for general business workflows, but manufacturing processes have a distinctive feature that such approaches ignore: the execution of every operation is accompanied by the consumption of a series of components and changes in their states. Two processes may look behaviorally identical on paper while consuming entirely different parts, or use identical parts while ordering their operations in critically different ways. When screening large process libraries for similar models, these blind spots cause numerous processes to be mistakenly flagged as highly similar, inflating review workloads and obscuring the real differences that matter for quality control.

To address this, the team first reformulated manufacturing processes as Petri net models, a formalism in which places represent products such as raw materials, items in progress or finished goods, transitions represent operations, and arcs capture the flow between them. Crucially, the model includes a function that assigns sets of components to each operation. Building on this representation, the authors introduce the concept of an operation profile, which characterizes how pairs of operations relate when a process executes. Five relations are distinguished: a strict relation in which one operation always precedes another, an interleaving relation that splits into circulation and parallel subtypes depending on whether the operations loop back to each other, and an exclusive relation in which two operations can never occur in the same execution.

A motivating example in the paper shows why a single criterion fails. When only operation profiles are compared, two processes with identical ordering relations score a similarity of exactly 1.0, implying equivalence, even though the components attached to their operations are not the same. Conversely, two processes sharing the same operation items but differing in whether two operations run strictly in sequence or in parallel are also hard to separate under a parts-only view. The lesson is that neither operation items nor operation profiles alone can deliver reliable judgments; both must be weighed together.

From this insight the researchers derive three similarity measurement methods. The operation items, or OI-based, method scores the similarity of corresponding operations by the Jaccard coefficient of their component sets, capturing how much the parts consumed by matching operations overlap. The operation profile, or OP-based, method instead counts how many operation pairs share the same occurrence relation, normalized by the number of possible pairs. The centerpiece, however, is the operation-aware profile, or COP-based, method, which fuses the two perspectives: for every operation pair it multiplies the pair’s component similarity by a binary indicator of whether the occurrence relation matches, then aggregates across the whole process. In a worked example, two processes that scored a perfect 1.0 under profile-only analysis drop to 0.87 once component differences are counted, and processes differing in both dimensions fall to 0.84, exactly the kind of downward correction that prevents false positives in process screening.

The same machinery extends to difference analysis. Given a newly created process, the method first retrieves all processes in the library whose COP similarity exceeds a threshold, set at 0.8 in the study. The operation profile matrices of these similar variants are aggregated into a single weighted matrix whose entries record the frequency of each occurrence relation across the variant set. Two quantitative indicators then drive an iterative clustering procedure. A separation value, computed with a weighted cosine measure, identifies which operations are most suitable to merge together, while a deviation value, measured against benchmark vectors for each relation type, determines which occurrence relation best characterizes the merged pair. Because different relations carry different constraint strength, from the weakly constrained interleaving relation to the strongly constrained exclusive relation, the weights are calibrated by combining domain rules with data-driven optimization rather than manual assignment. The result is a candidate operation profile matrix that summarizes the typical behavior of the whole variant family, against which the new process is compared entry by entry to expose every structural difference.

To test the approach, the team assembled a dataset of low-cost rapid manufacturing temperature sensors, establishing 12 reference processes with 4 variants each, drawn largely from the published literature and converted into Petri net form. A benchmark for true similarity was set with the help of eight experienced modelers, and the methods were evaluated on precision, recall, F1-score and accuracy. The COP method achieved a perfect precision of 1.0 in every case where similar variants existed, correctly rejecting dissimilar processes that the single-criterion methods accepted. The OP method, by contrast, reached recall of 1.0 in most cases, but its broad matching strategy came at the cost of substantial noise, with accuracy in some cases falling as low as 0.25. The COP method attained an average accuracy of 0.83 while the older OTV approach, which considers only part of the profile information, trailed behind in precision.

Statistical validation reinforced the picture. Paired t-tests at a significance level of 0.05 showed that the COP method’s improvements over the OI and OP methods were significant across both precision and accuracy, with all p-values below 0.05. Against the advanced OTV method, COP retained a statistically significant advantage in precision, while the difference in accuracy was not significant, with a p-value of approximately 0.1039, indicating at least comparable overall performance. The authors note that COP nonetheless offers greater theoretical completeness and interpretability, since it fully integrates both operation-component relations and the complete set of occurrence relations rather than a subset.

Perhaps the most practically significant finding concerns the method’s strictness toward small deviations. Rapid manufacturing processes are sensitive to environmental factors, and minor non-standard changes in operation sequence or execution logic frequently arise in production. The t-test results confirmed that when such substandard structural changes occur, the COP method assigns markedly lower similarity scores, flagging potential quality risks that looser methods gloss over. The authors therefore position the OI method for production resource arrangement and equipment scheduling, the OP method for macroscopic process planning and template construction, and the COP method for compliance verification and production quality control, where catching hidden deviations matters most.

In terms of difference mining, simulation experiments on grouped process models demonstrated that the iterative aggregation and clustering procedure correctly recovers candidate operation profiles, even when one process variant dominates the population with a 70 percent share, and that comparing candidate operation items against a new process pinpoints exactly which components deviate from established practice. The work was supported by the National Natural Science Foundation of China and several Anhui provincial programs. The authors acknowledge that, constrained by the availability of real industrial data and privacy limits, the evaluation relied on benchmark datasets from the literature; future work will apply the model to actual production logs with noise in real manufacturing workshops. If those trials succeed, factories could gain a rigorous, quantitative tool for managing sprawling process libraries, promoting reuse, and standardizing production on a scale that manual review has never been able to match.

Subject of Research: A Petri net-based method for measuring similarity and detecting differences between manufacturing processes by jointly analyzing operation profiles and operation-component relations.

Article Title: Analysis of similarity and difference in manufacturing processes based on petri nets

Article References: Analysis of similarity and difference in manufacturing processes based on petri nets. (n.d.). https://doi.org/10.1007/s10586-026-06494-y

Image Credits: AI Generated

DOI: 10.1007/s10586-026-06494-y

Keywords: Petri nets, manufacturing processes, similarity measurement, operation profile, process mining, difference analysis, behavioral profiles, process optimization, Cluster Computing, t-test, rapid manufacturing, precision and recall