by Jacek Korneluk

After spending half of my life at a bench, or in my laboratory office I acquired a good understanding of how this pathology machinery works. It is worth mentioning that my primary interests were the quality of results and process improvement rather than the “Business of Pathology” itself.  In 1994 I established a private pathology named Spektrum ZDL, same time still managing a public Core Lab. Then after 13 years of private pathology experiences, I decide to move from Poland to Australia.

In my entire laboratory career, I used all sorts of techniques and laboratory methods. A manual method, usually laborious and sometimes a bit “ancient” however still considered as “referenced” and Gold Standard. Then all sorts of semi-automated, automated systems, including gigantic analysers setups and laboratory robots. All in the name of doing more, quicker, and in the spirit of so-called high throughput.

Nowadays commercial laboratories trying to accumulate and run as many samples as possible, very often under-stuffing human personnel. A gigantic variety of all sorts of laboratory machinery and analysers form a structure of the system named high throughput. It is designed to constantly proceed, analyse laboratory samples, and produce test results. As quick and efficient as possible, and the more the better.

It is working but there are some drawbacks, especially regarding pathology workforces. Constant push and growing expectations combined with multitasking affect the performance of laboratory personnel which in consequence may not be consistent with high quality and good laboratory practice (GLP).

High throughput does not need to mean a cutting thought working laboratory environment.

Today we have AI with AI- Agents, and blockchain. Both technologies are friendly to all sorts of labs IMO. I mean not only to Pathology Services but to the entire Research, Digital Health, Medical Administration, Health Funds, Etc.

Despite still being considered as nascent and in the early stages a lot of Medical-AI applications are ready to be fine-tuned and deployed. Usually, they are separate platforms, fragmented, and not capable of making a complex system. One system for multiple users and with multiple entries for example.

Or a few systems that can cooperate and communicate together, seamlessly especially in a laboratory diagnostic environment.

There are plenty of publications and existing platforms introducing you to Digital Health and Digital Pathology. Some of them are imperfect, others very professional. They are usually waiting to be rediscovered, redeveloped, approved, and adopted.

On top of creating my own pathology, I was co-designing an alternative Medical System in Poland, based on a private ownership model. In the late 90s, I made my contribution to the creation of privately owned outpatient’ clinics which helped to reshape health care delivery. It was considered an innovation and the clinics doing well by the time of writing.

I started with Transfusion Medicine by proposing “putting” Blood Bank on a blockchain.

Later designing a digital pathology policy proposal titled “Diagnose with care – Primum non Nocere”.

Finally, focusing on Developing an AI Strategy for Modern Pathology.

I received good marks, and I am proud of it.

Recently, I was invited to a panel at a Digital Transformation meetup where we were focusing on AI for Excel and Power BI solutions. It inspired me to think deeper about pathology setup. Below are some facts I discovered while doing my own research (DYOR).

Clinical labs handle a gigantic volume of tests daily, requiring quick and accurate results.

Proposed use of tools:

Excel-based tool for recording and organising test results.

Power BI to monitor lab performance metrics like test turnaround times and accuracy rates.

AI Integration: AI algorithms analyse test results, flagging abnormal outcomes for further review.

Pathology Copilot: To further integrate, improve, and streamline internal and external processes and make life of pathology workforces easier.

Automate Test Result Analysis and Reporting in Clinical Laboratory Diagnostics settings to the new capacity and working dimension.

Overcome the risk of overloading, and difficulties to scale.

Excel

Role: Excel is used as a primary tool for recording and organising the vast array of test results. This includes patient identifiers, types of tests conducted, results, dates, and any additional notes.

Advantages: Excel’s familiar interface and powerful data organisation capabilities make it an ideal choice for maintaining structured records. Functions and formulas can be used for preliminary analysis and sorting. Also, it is an effective front-end for interacting with blockchain data and systems.

Power BI

Role: Power BI comes into play for more advanced analysis and visualisation. It’s used to create dashboards that monitor various lab performance metrics.

Test Turnaround Times: The time taken from receiving a sample to providing the test result. This is crucial for labs aiming to improve their efficiency.

Accuracy Rates: The rate of error or discrepancy in test results. Monitoring this ensures the reliability of the lab’s work.

Volume of Tests Processed: Keeping track of the number and types of tests helps in resource allocation and management.

Real-time feedback: Power BI’s real-time data processing and interactive dashboards allow lab managers to quickly assess performance and make informed decisions.

Role: AI algorithms, particularly those based on machine learning, are integrated to analyse test results.

Functionality

Flagging Abnormal Results: The AI system is trained to recognise patterns in the data, identifying results that deviate from the norm and flagging them for human review.

Calibration and Quality Management: automate the process of everyday QC control and quality assurance (QA).

Predictive Maintenance: Predicting when equipment needs maintenance before it fails, based on usage and error rates.

Advantages: AI enhances the lab’s capability to handle large datasets, reducing the manual workload and the potential for human error. It also speeds up the process of identifying critical anomalies in test results and automates test repeating, reporting, or adding.

Diagnostics Increased Efficiency – Automation and AI analysis streamline the workflow, allowing labs to process more tests faster.

Improved Accuracy – AI’s capability to detect anomalies and Power BI’s analytics reduce the risk of human error, leading to more reliable test results.

Enhanced Patient Care – Quicker, more accurate test results mean faster diagnosis and treatment for patients, directly impacting patient outcomes.

Data-Driven Decisions – The insights gained from Power BI dashboards aid in making strategic decisions, from staffing needs to equipment upgrades.

Improved Security and Privacy – If Blockchain technology is used.

With lives on the line, the pressure to process such capacity efficiently and reliably is immense. Unfortunately, relying solely on human effort has proven inadequate.

But now, labs may turn to technological tools like Excel, and Power BI combined with artificial intelligence (AI) and Pathology Copilot to take their operations to a higher efficiency level. Using together smart analytics and automation can significantly boost almost all modern pathology aspects. No more pushing bounders and pathology burnouts. Thank you, AI and Blockchain :).

Excel plays a key role in recording and organising the flood of test data. It may be also the gate to the blockchain.

Power BI takes it further with real-time dashboards that track key performance and analytical metrics. Interactive visualisations provide real time monitoring of turnaround times, accuracy rates, test volumes, and additional criteria. This arms senior scientists and lab managers with actionable insights to drive continuous improvement and informed decisions.

AI algorithms can rapidly analyse results as they come in, recognising patterns and spotting anomalies. This takes a huge burden off scientists’ shoulders, reducing the chance of oversight and human error. It may help work at the validation desk or QC stands to be less stressful and more efficient. Also, eliminate to minimum human error. The Pathology Copilot is always ready to help!

From flagging abnormal test results, validating QC, and reporting extremes, to automatically ordering additional analyses, or repeating inconsistent test results, AI quickly identifies critical cases needing further analysis, accelerating the diagnostic process in a fully automated manner.

No more accidents of reporting unreliable outcomes without confirming and validating QC first.

AI also enables Predictive Maintenance, forecasting equipment issues before they lead to failures. This boosts uptime and reliability.  The Automatic Inventory Management and Ordering is another good example. The list is endless as the needs of Pathology are.

Finally, the proposed setup enables real-time delivery of results and query on the move! For doctors and nurses when equipped with “Apps in the field”.

AI-data-driven decisions combined with Power BI insights could help optimise staffing, equipment, budgets, and more.

The impact on workload, efficiency, accuracy, and patient care is transformative.

What once seemed like routine automated laboratory workflow is now smartly turbocharged by intelligent tools. Test results delivery that used to take days can now be turned around in minutes or hours.

Abnormal results get escalated instantly before they’re missed or wrongly reported. And managers can adjust operations on the fly based on real-time analytics.

The technology is definitely here and now– we just need to be open-minded, and innovative enough and go for it!

Cheers, Author.

PS.

I would like to extend my sincerest appreciation to Professor Adam Danek, Professor Leon Liszka, and Professor Jerzy W. Naskalski for their invaluable guidance and expertise throughout my academic journey. Their dedication, insight, and support have significantly contributed to my development and achievements. I am deeply grateful for their mentorship.

Originally published at

https://spektrumlab.io/revolutionizing-pathology-ai-blockchain-and-analytics-integration/