As reported in a recent edition of the Harvard Gazette, AI is commonly used to analyze X-rays and scans. AI has begun to outperform physicians in prognostic and diagnostic assessments.
AI can analyze medical images to identify suspicious areas in scans like mammograms, CT scans, or MRIs, potentially detecting cancers earlier.
In addition, technology has been raised to a higher level with the introduction of large-language AI models (LLMs). A simple description of LLMs is that they are computer programs that answer questions and can participate in conversations.
The director of Harvard’s Center for Bioethics, Rebecca Weintraub Brendel, emphasizes that knowing that “we can” does not give license to “we should”.
In one respect, either way there is a similarity in that ultimately the patient’s wishes become the last and final decision. That is providing that the patient is competent to make the decisions and their wishes constitute something that is medically indicated.
The patient may also be so ill that they are unable to express their wishes.
The Effects of Depression
Depression is quite common in certain types of cancers such as lung, breast and pancreatic cancer. It can be severe. Young patients are likely to experience depression as well as people who have experienced depression in the past or have a higher tumor cell burden.
Note that some patients change their mind about their end-of-life choice after receiving treatment and their symptoms are alleviated.
This also applies to devastating injuries such as spinal cord injuries where patients will report experiencing a better quality of life. Therefore, a person’s capacity to change must be considered when making end of life decisions.
If the patient has a chronic illness, AI would be helpful by making end-of-life decisions at various times during the illness. However, at the onset of the illness the patient may want to be presented with a treatment plan or palliative care.
In these situations, AI might forecast the tipping point for a specific patient. AI’s ability to process more information than that of the human mind may be helpful. Especially since that information will not be influenced by anxiety or commitments.
The end-of-life decision falls on the care team in cases where a patient is incapacitated, without a family in attendance and has not prepared an advanced directive.
The team can only assume that these patients would not have requested treatment. In such instances AI would be helpful.
Artificial Intelligence (AI) vs the Human Touch
Is it morally acceptable for AI to make end-of-life decisions versus just gathering information? What guidelines should be in place with respect to human performance?
Dr. Brendel suggested that when a patient made it clear that he or she does not want to be resuscitated, unless the patient tells the doctor otherwise, he or she will not be resuscitated unless it is determined that there is a good chance of recovery.
Some question whether people should always be involved in life events.
Should the first contact in this world for a baby be human hands or should we focus entirely on quality of life?
What responsibilities do doctors and patients have to each other.
How do we continue to give meaning to important life events.
And in conclusion, how do we educate and make sure the healing professions will prioritize delivering care?
Source: The Harvard Gazette
