Machine Translation systems have become valuable tools for everyday tasks. Their application in more robust contexts, like scientific research, is still an ongoing endeavor, specifically in study domains where the outcomes of these systems can potentially impact the life of a human being. With advancements in Natural Language Processing literature, newer translation models, capable of exploiting scarce amounts of information, can be used to create, expand, or adapt corpora needed to assess critical phenomena, such as Suicidal Ideation Detection, in low-resource languages. The present study aims to compare the performance of novel Machine Translation techniques based on standardized metric evaluations, particularly in translating suicide-related language from English to Spanish. Further, this research aims to determine which of these methods performs the best when used in life-threatening-related data, making it one of the first works to emphasize such implications using this set of languages.