Digital transformation: Applications in the laboratory and in-vitro diagnostics

It has been said that we are in the Fourth Industrial Revolution. The first in the 18th century was powered by steam, the second in the 19th century was made possible by electricity, and the third in the 20th century saw the beginnings of digital automation. Today, we are in an era of digital revolution.¹

We are seeing disruptive technologies that could completely change the way businesses are run, such as advanced automation and connectivity driven by artificial intelligence (AI) and machine learning.^2,3 At Roche Experience Days 2024, Dr. Deng Kun, Director of Laboratory Medicine Center at Third Affiliated Hospital of Chongqing Medical University, China, spoke about what digitalization means for laboratories.

A recent book by McKinsey describes digital transformation as ‘the fundamental rewiring of how an organization operates’ and Dr. Deng is clear on the importance of embracing it.³ “Digital transformation reshapes companies through the strategic integration of digital technologies, leading to fundamental changes in how businesses operate and deliver value to customers. For us, that is patients,” says Deng. “Digital transformation is not a choice, but a necessity in today’s world.”

Dr. Deng notes there are many digital technologies available to labs to enable digital transformation. These include the Internet of Things, robotics, AI, machine learning, cloud computing, and big data. Deng goes on to explain that these technologies “come together to form what we call digital ecosystems. The digital ecosystems work together to shape and modernize the way we operate.”

Achieving a digital ecosystem involves a cyclical process, which Dr. Deng outlines. It starts with digitization, where physical objects are stored in a digital format.⁴ For example, scanning historical pathology slides and saving them in a digital database. Then comes informatization, which is the introduction of information solutions such as a Laboratory Information System to optimize the processing, storing, and managing of data.^5,6 Next, processes and workflows are automated with software solutions where possible to minimize the need for human intervention.⁷ Lastly, digitalization looks at all the digital information available, and this can then be utilized to support decision-making or improve processes in the lab.⁴

For Dr. Deng’s lab, this was followed by three key steps:

• Connecting everything to enable seamless data flows
• Automating end-to-end workflows to enhance efficiency
• Utilizing advanced analytics to provide actionable insights

In taking this approach, the team transformed its lab into a fully integrated and intelligent digital assessment ecosystem. Dr. Deng is keen to point out that the goal of building such an ecosystem is not to have something that ‘appears grand and sophisticated,” but rather that “We need to deliver real benefits for patients by ensuring testing availability, improved accuracy, better clinical decision making, and faster turnaround times.”

In the hospital laboratory, Dr. Deng and colleagues have implemented several successful digital applications across their ecosystem.

The team kick-started its digital transformation journey with automation. “We took immediate action by equipping our laboratory with state-of-the-art automation systems,” Deng explains, “and this significantly enhanced accuracy, reduced turnaround times, and optimized operational efficiency. By automating repetitive processes, we not only achieved cost savings but increased our revenue potential.”

“Automation and quality control can work hand in hand,” says Dr. Deng. “By implementing automatic quality control systems, we can set up an automated schedule to perform quality control and test automatically before the official start of the workday. So, it's not only efficient but also gives our employees more sleep and makes for happy staff!”

An advanced approach to quality control is patient-based quality control. Dr. Deng explains that by collecting patient data over six months or a year, assay testing parameters can then be simulated in an algorithm. These are then configured, verified, and confirmed as parameters allowing for the evaluation of assay performance using data specific to the laboratory.⁸

Another crucial application, according to Dr. Deng, is intelligent serum quality assessment. This technique involves using advanced algorithms such as You Only Look Once (YOLO) and the Convolutional Neural Network (CNN) classification algorithm to analyze images of samples and identify quality issues such as hemolysis, lipid content, and icterus. “The system gave us high-quality results while minimizing manual errors, reducing costs, and optimizing turnaround time to enhance reliability.”

Across digital transformation efforts, Dr. Deng points out that it is important to keep in mind meeting standards for diagnostic operations, such as the recently updated International Standard ISO 15189, but also notes that digital transformation is not just about operational improvements, but that it is also driving innovations in diagnostics.

A promising area in diagnostics is the development of digital biomarkers. By combining traditional biomarkers with biomarkers collected using digital devices such as smartwatches, for example, a patient’s heart rate and physical activity level, Dr. Deng believes there is a potential to “revolutionize the detection and monitoring of diseases and provide accessible medical support for patients and doctors.”

An example of a successful combination in the laboratory is the gender (biological sex), age, alpha-fetoprotein (AFP), and des-gamma carboxy-prothrombin (GAAD) algorithm for earlier detection of liver cancer in patients with chronic liver disease. This algorithm integrates key indicators such as gender, age, AFP, and bipartite levels, and data is analyzed to generate a GAAD score, which indicates the likelihood of the patient having cancer. The algorithm was developed and validated across many countries with diverse cohorts. “It demonstrates robust diagnostic capability that surpasses existing biomarkers,” says Dr. Deng, “This comprehensive process from algorithm development to clinical validation represents a major step forward in precision medicine for patients.” 

Dr. Deng believes digitalization has made a profound impact already but reminds us that this is just beginning: “For all of us, digital transformation is a continuous journey. In the journey of digital transformation in in vitro diagnostics, we need to be clear that the needs of the patient remain at the heart of our efforts.”

For further details on the technologies enabling digital transformation in the laboratory, watch Dr. Deng’s full presentation here.