How AI solves depression and finds cancer by Suf Alkhaldi

AI will make our life much easier
AI will make our life much easier

Artificial Intelligence AI moves us in different fields of life, faster than we thought even two years ago.  Using AI in diagnosing bipolar depression as well as other illnesses has created a buzz in scientific communities so that many startups have started to use it.  One of these companies is NeuroLex Diagnostics Inc.  The company has developed a powerful natural language machine learning processing to not only diagnose schizophrenia and bipolar depression, but also to monitor patients’ language while they are under treatment to determine if the medicine is working and whether the doctor needs to change the medicine or the dose.

It turns out that mental illness manifests itself in patients in their use of the semantic syntax of language.  People who are depressed use speech patterns which psychiatrists know very well. In the future, recording patients’ conversations and analyzing them by an AI machine during a doctor’s visit may become routine like taking a blood sample. This AI revolution will definitely decrease the cost of treatment by diagnosing depression, schizophrenia, and other common illnesses much faster and cheaper.

As this idea is still being tested in several hospitals, including Massachusetts General Hospital, where psychiatrist  Dr. Arshya Vahabzadeh uses AI in speech recognition, the machine learning research and its application will open the door for other fields, including crime prevention, anthropology, as well as love!

AI is currently being used at Stanford University to develop a machine learning algorithm to identify malignant lung tumors that far exceeds human ability to read tissue slides. Stanford professor Dr. Michael Snyder points out that reading tissue slides to identify cancerous tumors is highly subjective: “Two highly skilled pathologists assessing the same slide will agree only about 60% of the time.”¹ 

The new AI was trained on over 2,000 slides, and it was able to diagnose cancerous tissue correctly in 10,000 individuals.  This achievement surpassed human pathologists’ ability which only were able to identity a few hundred.   Clearly, AI does not have any biases and fatigue–the usual culprits that interfere in the performance of human pathologists . Do we still need doctors to read these slides or X-ray films?  Definitely, AI will be another layer of competence to advance the performance of the diagnostic process.

What does this mean for the future of AI in medicine? Many studies have revealed that computer algorithms are much better at drawing correlation findings.  As AI becomes better and better, the ability of these AI codes to adapt to many challenges, from reading an x-ray film to analyzing speeches of people, will advance incredibly.

The capability of AI, in fact, cannot be imagined right now.  AI will be breaking or changing the definition of many careers, especially doctors.  It will break doctors’ jobs into sub-tasks and slowly but surely computerize the tasks, eliminating the need of human input. The outcome of this move will increase human productivity and reproducibility, moving the benefits to a lot of emerging markets. Even doctors’ jobs, unlike what many doctors believe, are not safe from being automated.  For all the people who are moving  forward in their career lives right now: Please adapt to the new AI revolution and learn as much as possible to use it before you can be replaced by it!

Thank you for reading my post. Please let me hear from you. (Sufalkhaldi@futureandsciencehacks.com).  This post is usually published on Saturday.

Leave a Reply