In August 2021, the news about Japan suspending 1.63 million doses of the Moderna vaccine shocked the world community. It was revealed that the vaccine was contaminated with metal particles due to “human error specific to visually misjudging the required 1mm gap between the star-wheel and the stopper” of the machine that put the tops on vials. This incident put the spotlight on vaccine inspection, and how it could be carried out at scale.

Clearly, new systems were needed. This mistake could have posed a threat to the lives of 1.63 million people getting their vaccination. 

New inspection methods needed

The pharmaceutical products are getting more complex, meaning that inspecting them is getting more complicated as well. This is reaching the limits of what is possible with traditional algorithms or manual visual inspection. The present pandemic situation, in particular, has made the whole global population dependent on the quality of COVID-19 vaccines. And it is a fact that biotech products are even more sensitive than other chemicals, which also results in more sophisticated packaging.

As data collection techniques become more efficient and widely available, many pharmaceutical companies are starting to consider Artificial Intelligence (AI) for improving their drug and vaccine inspections. Machine Learning (ML), a subsection of AI, takes advantage of this improved data collection to create algorithms that can detect defects in real-time with human-level accuracy.  For this purpose, ML engineers use supervised and unsupervised algorithms. Unsupervised algorithms detect very rare things such as mixed-up products, whereas supervised ones identify small objects like foreign body particles, pieces of glass or metal, hairs, etc. 

Current inspection approaches: why AI is better

Currently, pharmaceutical companies do drug and vaccine inspections using two approaches: manual visual inspection by human operators and classic Computer Vision (CV) algorithms. AI algorithms should be considered instead of manual inspection because they eliminate the chance of human error, thus providing more consistent and accurate results. At the same time, since using these types of algorithms reduces the number of workers involved in inspection, the remaining workers can focus on other tasks, which also secures labor costs for the company. 

In comparison to classic algorithms, ML algorithms represent a more advanced inspection technology. They are not programmed beforehand to do a task but instead learn how to do it themselves by analyzing their own performance and learning from mistakes. However, once the algorithm gets the best metrics possible, ML engineers can freeze it to reach the quality goals.

In addition, classic algorithms are often not feasible for an environment that is in constant change. They employ a more rigorous mathematical approach, whereas ML algorithms can be applied to any situation where there is available data and are therefore more suitable for visual inspections. AI algorithms process data in raw form which, in the case of drug inspection, includes images. 

Corporate innovation: easier said than done

Interestingly, the number of companies actually implementing AI in inspections is drastically smaller than of those willing to apply it. This is often explained by the absence of teams with cross-functional skills at a corporate workplace. For creating a well-functioning AI process, it is necessary to have a team with people possessing both the knowledge of IT development and machine learning science as well as quality inspection and pharmaceutical regulations, for instance. This frequently poses a problem in the corporate environment where every individual or department has a specific set of skills. 

Consequently, the solution cannot be implemented without consulting with the whole corporation. Normally, in this case, the company needs to build an entire innovation department from scratch with people of relevant skills, which often takes from 3 to 5 years. They also require material to work with. Therefore, companies also need to invest in the development of special hardware such as robots, cameras, servers, and screens for the interface to interact with robots or robotic arms. 

This is why tech companies that provide AI for drug inspections emerge. It is true that the teams of small companies are much more multifunctional simply because their size forces them to be so. Despite all the bureaucratic drawbacks, tech innovators do set up pilot projects with pharma giants to demonstrate how easy it is to implement AI. This often simplifies the change management process at the corporation because they can co-work with a team of a small company that already has a solution for them and can help them with introducing it at the workplace. As a result of such cooperation, pharma manufacturers endure fewer false rejects, deviations, and recalls. 

Challenges on hand

Currently, these tech companies face two main challenges. The first one is improving the algorithms to reach 100% accuracy. It is proved that the visual inspections with AI reach up to 90% precision, demonstrating some room for improvement. The second challenge is overcoming the barriers to entry in the pharmaceutical environment. All the new technologies need to comply with FDA and Eudralex regulations to be used in the pharmaceutical environment. 

Eudralex Volume 4 Annex 11: Computerised systems provide a perfect example. According to this directive, the manufacturer of the medical device which is going to be implemented in the pharmaceutical industry must guarantee proper validation, electronic archiving and signatures, risk management, and security. Only this regulation alone requires an extra effort from AI developers while they also need to ensure a high level of security considering the confidentiality of corporate data and regularly validate the computer system. Considering that the AI algorithms are a “blackbox” by essence, validating and controllblack boxthe process steps might be sometimes challenging.

Postponing change has fatal consequences

The pandemic has clearly demonstrated the necessity for a balance between having strict regulations and proper change management in pharmaceutical companies. Even with all the pressure and time constraints, it took more than 1.5 years for the most advanced enterprises in the field to actually start the mass production of COVID-19 vaccines. This is clear evidence of pharma manufacturers lacking flexibility. 

The innovative approaches for drug manufacturing like AI, for example, have previously been neglected. Most pharmaceutical companies put “extending human lives” and “improving the quality of global healthcare” in their mission statements. This, of course, explains why there are so many standards and procedures in place: pharmaceutical companies cannot risk the life of patients. However, in a state of emergency, when the destiny of the whole global population is at stake, it is the duty of these companies to deliver a result as soon as possible at any cost and measure, especially when the innovative solutions are here for them to use. Refraining from doing so simply contradicts their corporate values.

Finding the right AI solution

And now the main question comes: how to find the provider of the best AI service for a pharma business? Since the concept of visual inspection with AI is very new, this is exactly why startups are working on the development of the relevant algorithms. It is, therefore, recommended to look for those startups that have references from pharmaceutical companies. Usually, the young companies working in this area do not have the legacy of developing other solutions. This means that the firms use their tech solution in full capacity: working on and improving exactly the few most important features for drug inspection instead of taking other use cases.

Clinicians pivoted their AI efforts to engage in the battle against the COVID-19 pandemic. How can it help those with persistent symptoms – the so-called “long COVID”?

The emergence in March 2020 of the COVID-19 virus as a pandemic had a profound effect on the demand for health services that continues to this day. The impact on laboratory services was particularly notable in the early days of the pandemic. A study helmed by Thomas J S Durant of the Department of Laboratory Medicine at the Yale University School of Medicine found that from late February to mid-April 2020, more than 870,000 COVID-19 tests were administered in the U.S., but overall lab testing went down significantly. Quickly, COVID had become a burden that was affecting other laboratory functions.

But some hospital laboratories had an ace up their sleeves: Artificial intelligence (AI). Some developed algorithms to predict the likelihood of a patient contracting COVID based on demographic data and vaccination history to prioritize then-limited testing resources. Some adapted existing projects to predict respiratory failure.

Some retrained radiological imaging AI projects were created to speed the diagnosis of thoracic ailments. Others used AI and machine learning (ML) modeling to find which patients needed less attentive care and those likely to require intubation (the procedure that’s used when you can’t breathe on your own). Massachusetts Institute of Technology (MIT) researchers also developed an AI model that could distinguish even asymptomatic COVID-19 sufferers with startling accuracy by analyzing recordings of coughs collected over mobile phones. The model was adapted from algorithms already proven to detect asthma and pneumonia accurately.

In short, AI and ML have eased healthcare burdens by de-escalating patients to free up resources, prioritizing testing for those most likely to have been exposed, and providing diagnostics with novel forms of data capture. While this benefits the predictive and early treatment phases of the COVID protocol, AI must also play a role in post-acute cases.

While most people recover from a COVID infection within a few weeks, some have symptoms that linger. Long COVID – or post-acute COVID, or chronic COVID – is defined as symptoms lingering for 12 weeks or more after infection that can’t be explained by another existing or recently acquired condition. The list of symptoms is long: trouble breathing, headaches, fever, “brain fog,” heart palpitations, joint or muscle pain, and changes in smell or taste are just a few. Long COVID can result from the initial infection’s damage to the lungs, the heart, kidneys, skin, and brain – practically any organ in the human body. And Long COVID isn’t just a problem for those who were severely ill or hospitalized because of COVID; it can manifest in COVID patients who were asymptomatic during the acute phase of the infection.

We can leverage the front-line AI and ML technologies to help manage long COVID suffering. Predictive models could ascertain the likelihood a patient could suffer from post-acute COVID, determine which patients are suffering from non-infectious consequences of COVID, like isolation and loss of income, and comb through the chemical composition of an entire library of pharmaceutical treatments that may be effective at treating post-COVID symptoms.

We have an enormous body of data to train artificial intelligence and machine learning platforms. But there is an impediment to its success: siloing.

When COVID-19 was first declared a pandemic, health care providers had to scramble to adopt or adapt AI tools to address their specific challenges. This meant that modeling and algorithm development took place in-house rather than in a shared forum. This leads to no guarantee of consistent data collection or output format that could be shared with other providers, even within the same region, where shared results would be most beneficial.

Interoperation of disparate systems is a vital component of a comprehensive approach to predicting, detecting, and treating COVID-19 and its post-acute complications. This requires standardization of reporting formats on a healthcare authority or statewide basis, ideally compliant with Centers for Disease Control (CDC) protocols. Data collection procedures must be formalized, digitized in a form that is resistant to error, and brought as close to the patient as possible for maximum accuracy.

AI and ML technologies offer nearly boundless opportunities in the life sciences field—modeling potential disease hot spots, discerning the desperately ill who require attentive treatment from those with a mild illness that can be self-managed, and predicting what resources will be needed and where to battle the pandemic. We have a head start on fighting Long COVID and the related symptoms by preparing for interoperability.

In 2020, the world experienced an unprecedented increase in cybercrimes amid COVID-19. In fact, data breaches increase 273 percent in the first quarter, compared to 2019, according to a new study from cloud computing company Iomart.

Thanks to the additional vulnerabilities that opened up as people work from home, moves to take everything digital and conduct all business online, and the general confusion caused by the pandemic, cybercriminals have taken full advantage of the situation.

That’s a significant problem on its own, but there’s another issue at hand that makes the situation even worse.

According to a report by (ISC)^2, the number of unfilled cybersecurity positions now stands at 4.07 million, up from 2.93 million this time last year. This includes 561,000 in North America.

The shortage of skilled workers in the industry in Europe has soared by more than 100 percent over the same period, from 142,000 to 291,000.

The report suggests a number of remedies for this situation, including in-house training, bringing employees across from other IT areas and retraining them, and increasing efforts to hire with aptitude in order to bring them up to speed on cybersecurity quickly.

One company has been helping to plug this gap. Cybint – a global cyber education company – recently partnered with LCC International University, an American-style university with students from over 50 countries, to create the Cybint Bootcamp.

Cybint also recently partnered with Israel-based web data provider Webhose and threat protection platform IntSights to provide a more well-rounded learning experience for Cybint users. These companies are part of the company’s effort to join forces with leading cyber technologies, bolstering the tools at its disposal to further reskill the workforce and upskill the cybersecurity industry.

And that’s important, because the shortfall of talent in the cybersecurity industry, combined with the rapid growth in attacks and breaches, is going to need to be dealt with quickly.

“We like to compare the cybersecurity market to that of coding and computer programming a few decades ago,” Roy Zur, CEO and founder at Cybint, told me. “Many of the first pioneers in this field were self-taught or learned by doing, mainly because traditional higher education just hadn’t caught up yet and employers were looking for the skills. Fast-forward, there are coding bootcamps and academies dedicated to this field as an alternative to degree education. Cybersecurity is similar in the way that the demand exists, but the skilled individuals aren’t necessarily coming out of higher education, and if they are, their skills are not always practical or relevant to real life. We believe that there is a huge opportunity for cyber professionals to learn skills quickly and effectively through intensive career bootcamp that are focused on the most in-demand job roles in cybersecurity.”

Security firm McAfee estimates the cost of cybercrime in 2020 reached $1 trillion, a figure that includes both the losses incurred and the amount of money spent on cybersecurity. If businesses are going to get a handle on these costs – which represent a 50 percent increase on 2018 – they are going to have to move fast.

So how long does someone have to train in cyber to become effective and gain employment in the field?

“Traditionally, it’s a matter of going through college and certification,” Zur said. “Alternatively, it could be as quick as three months in the full-time Cybint Cybersecurity Bootcamp. My extensive background in cybersecurity military training mixed with my CPO’s background in building career boot camps at MIT has allowed us to put together a learning experience that is incomparable to what’s currently available. It’s practical, highly-focused, and interactive – exactly the experience that employers are looking for in their candidates.”

That focus on getting students from starting the course to being employable is important to Cybint, and crucial for businesses everywhere.

“We are truly career-focused,” Zur said. “Our end goal is to help our Bootcampers land high-paying and long-term opportunities in the market. We’ve tailored the Cybint Bootcamp and our business model to achieve this to ultimately close the workforce and skills shortage in cybersecurity”

So what’s next for Cybint?

“There are quite a few avenues we can take as we scale,” Zur said. “However we plan to stay true to our mission of tackling the workforce shortage and skills gap through skills learning and collaboration. With that said, we plan to offer the Cybint Bootcamp in more locations worldwide through our partners and expand the cybersecurity roles we train for.”

One thing is certain. With such a huge increase in cybersecurity attacks, and the huge skill gap we’re currently experiencing, 2021 is already set to cost organizations as much as it did in 2020. Those willing to move across to cybersecurity can see this as an opportunity – the cyber market is forecasted to grow to $248.26 billion by 2023, making it a lucrative area, and one that may rival that of other high paid IT roles, such as data science, analysis, and engineering.

This article originally appeared at Grit Daily, and is reproduced with permission.