Why do neurologists revise AI algorithms designed by engineers?

Revolutions happen when humans and machines work in tandem. But why do we need neurologists to design AI algorithms, when we have engineers for that? The answer is to be free from bias, which cannot happen without the intervention of neuroscientists.

Scenario without neuro-experts

Artificial intelligence is evolving and the number of respondents implementing AI has steadily increased to almost 90%. Studies revealed that an average investment of $40 million has been recorded so far for the next five-year AI plan. The need to involve neurologists grew out of previous AI algorithms that were biased in terms of gender, race, and other variables. This bias has raised questions about the engineers designing the AI ​​models for deployment in clinical trials and across healthcare settings.

A report of Chilmarka global research and consulting firm providing IT solutions for health, has pointed the finger at discrimination freshman COVID-19 pandemic algorithms, which negated the accuracy of X-rays and CT scans. The most common bias comes from the imprecision of the algorithms which cannot define the target population purely designed by engineers.

AI algorithm designed under expert supervision

Emphasis should be placed on its design in such a way as to constantly help physicians update information from old or new published journals, articles or manuals in order to improve patient care. It simply means extracting important information from detailed patient data to calculate the probability of disease progression. And neuro-experts can pretty much understand the nuances of genetic data design and structural imaging, allowing for more clarity about the patient’s exact problem. Methods such as NLP further help extract unstructured data (medical journals) and turn it into machine-readable data, which MLs can then review, if properly trained under clinical supervision, not just in engineering design . Otherwise, it may have contradictory effects on the complete data analysis.

Neuro experts safeguard the future of AI imaging

The process of filtering images based on deep learning is now widely recognized by various scientists and experts. This algorithm helps to analyze strokes and cerebral hemorrhages through early stage images. If the computer detects a problem in their image analysis, their file will immediately move to the priority segment. And vice versa on priority list, based on image analysis detection algorithm. However, if not designed well, the AI ​​control over image quality, report structure, and computer-aided classification would suffer.

Such deep learning algorithms can improve the quality of the MRI image, which neuro-experts can design correctly, and thus ensure improved neuro-imaging technology, better preoperative assessment and network analysis. Resting-state fMRI grading of glioma, presurgical localization of the eloquent cortex, and epileptic focus are some examples of conditioned AI imaging. Additionally, experts designing AI can help sort through the stacked datasets produced by modern neuroscience tools, such as real-time imaging and multi-electrode arrays. Along with this, a correct AI algorithm can help integrate more knowledge discovery, data mining, segmentation, pattern recognition, graphical visualization and other crucial areas yet to be invented over time.

Nishit Agarwalan American engineer in biomedical data science, spoke with Analytics India Magazine on the need to bring clinical experts in neuroscience to design algorithms, thus not giving the reins to engineers. He currently works for Medidataand previously worked as a computational neuroscientist with expertise in creating intelligent algorithms for processing physiological data collected from the human body using various wearable sensors.

AIM: What do you think is accelerating the penetration of AI into healthcare?

Agarwal: The last decade has been quite unpredictable for the world of artificial intelligence and medicine. As more physicians become familiar with machine learning in healthcare, countless innovations in symptom tracking, disease diagnosis, and precision medicine are being developed. Post-pandemic, medical-grade wearable devices paved the way for clinical trials at home. Finally, AI and machine learning techniques have enabled neuroscience research to capture the underlying pattern of neural data, accelerating its importance everywhere. Today, startups like Neuralink, neural, next spirit, emotional and many others are investing millions in the development of AI algorithms.

OBJECTIVE: Why do we need clinical neuroscience experts in AI design rather than engineers?

Agarwal: Data lacks quality without the presence of clinical experts to design them. The algorithms depend on factual data, and the quality of this data directly depends on the experiments carried out and the quality of the electrodes. Unfortunately, until now, the contribution of experts to the design of these AI algorithms has been minimal, which shows a huge gap in knowledge of the field. Neuroscientists can greatly affect the quality of information gathered from data by helping to design holistic experiments. Their expertise removes barriers and biases for engineers in achieving the goal of creating algorithms. Thorough evaluation and constant oversight by clinical neuroscience experts could help data scientists like me change the way we interact with the world directly from our brains.

AIM: What is the future of this problem?

Agarwal: The use of AI in neuroscience is still in its infancy, but the NeuroAI community is growing rapidly. More and more engineers, data scientists and AI enthusiasts are coming together to solve the challenge of understanding the brain and using AI for breakthrough innovations in neuroscience. Hopefully change the perspective of its development and give control to clinical experts to shape AI algorithms for a bias-free world and use.

AIM: Do you think neurologists are shaping AI?

Agarwal: Neurologists are shaping AI to give better clarity to the human brain by mimicking its mental functions. Moreover, they are focused on discovering a massive amount of relevant information through AI.

Sharon D. Cole