The authors, Wijayanti et al. [1] of an article in this journal titled “Artificial intelligence implementation in the management of patients with tuberculosis” have proposed the development of a natural language processing (NLP) tool to detect depression in patients with tuberculosis (TB). They suggested that such an instrument would help in the early detection of depression, which could facilitate timely intervention and, in the long run, improve treatment outcomes.

NLP is an artificial intelligence (AI) tool that allows computers to understand human language, whether written, spoken, or even scribbled [2]. Recent studies suggest that AI tools, such as NLP, can analyze text and speech patterns to objectively measure depression. For example, DeSouza et al. [3] established that NLP approaches are a promising means to help assess, monitor, and detect depression and other medical conditions in older individuals based on speech. According to these findings, late-life depression can be detected using NLP. Another study highlighted NLP models such as BERT, Llama2-13B, GPT-3.5, and GPT-4 as potential tools for detecting depression [4].

The proposal to develop an NLP framework to detect depression in patients with TB would be undeniably beneficial for the immediate recognition of illnesses. However, ethical concerns may arise by solely relying on AI tools like this. Such reliance may overlook the need for personal encounters wherein showing empathy and compassion may occur. The subjective experiences of each individual are vital for effective treatment. The actual encounters between each person and the clinician provide an avenue for unique subjective experiences that influence their coping mechanisms and recovery. Direct observation by the clinician, personal judgment, and a nuanced understanding of the context are irreplaceable for truly addressing the condition. The actual communication and interaction with the patient are incomparable for better interpretation and diagnosis through the exploration of nonverbal cues and other details in the communication process. Alternatively, differences in diagnosis and treatment recommendations could arise owing to the bias of AI models.

Overall, AI tools should be regarded as complementary tools, and never as replacements for actual medical practitioners, whose empathy, contextual understanding, and personalized care shown to each person are irreplaceable.

Citations to this article as recorded by