How exactly did you determine that 452 articles were pertinent to RQ1, 1,191 to RQ2, and 140 to RQ3 before removing duplicates? A single database search using strings A, B, and C would return overlapping results. Your reported numbers show no overlap, how do you explain this?
Can you provide the actual search syntax used? The strings in Table 1 combine terms with AND and OR arbitrarily, and it’s unclear how these were actually executed. For example, string B contains ((((TS=(challenges)) AND TS=(limitations)) AND TS=(image processing)) OR TS=(AI techniques)) AND TS=(cancer), this retrieves everything with AI techniques and cancer regardless of whether challenges/limitations are mentioned. Your RQ2 is about challenges and limitations, yet this string captures articles that simply mention AI techniques and cancer without addressing challenges at all.
Were the 1,783 articles the result of running these three strings separately and then combining the results? If so, how did you handle duplicate articles retrieved by multiple strings before assigning them to a single RQ?
Without a transparent, logically consistent search process, the 57 articles selected for analysis cannot be considered a systematic representation of the literature. This is a foundational flaw that calls into question every finding presented.